Attempt at new dataloader, much more automated.

data/AIInitiatives/train.csv

@@ -1,66 +0,0 @@
-Name,Organization / Author,Brief Description,Sector,Geographical scope,Target Audience,Stage of Development,Date started,Country/region of origin,Notes (including specific SDG(s) and OECD AI Principles addressed),AI AND ETHICS (Ethical frameworks and guidelines promoting and fostering responsible AI),"AI AND GOVERNANCE (Governance mechanisms operationalizing responsible AI, including auditing mechanisms, risk assessments, standards, certifications, corporate governance frameworks, etc.)",AI AND SOCIAL GOOD (Applied projects advancing SDGs responsibly)
-Limits on autonomy in weapon systems. Identifying practical elements of human control,"Stockholm International Peace Research Institute, International Committee of the Red Cross.","There is wide recognition that the need to preserve human control over weapon systems and the use of force in armed conflict will require limits on autonomous weapon systems (AWS). This report from the Stockholm International Peace Research Institute and the International Committee of the Red Cross offers in-depth analysis of the type and degree of human control that is required to mitigate the risks posed by AWS. It proposes three types of control measures to reduce or compensate for the unpredictability introduced by AWS and associated risks for civilians: controls on the weapon’s parameters such as types of targets, controls on the environment of use and controls in the form of human supervision. Limits on Autonomy in Weapon Systems: Identifying Practical Elements of Human Control is a comprehensive examination of the specific controls on AWS needed to ensure human control over the use force, and to address legal, ethical and operational concerns. It provides policymakers with practical guidance on how these control measures should form the basis of internationally agreed limits on AWS—whether rules, standards or best practices.",Academia and International Organisation,Global,"Policy makers, diplomats, civil society, international organizations",Published report,2020,Sweden,"SDG 16 Peace, justice, strong institutions",Yes,No,No
-”Ethics of Artificial Intelligence and Robotics” (Stanford Encyclopedia of Philosophy),Vincent C. Müller,"Encyclopedic entry in Stanford Encyclopedia of Philosophy. It covers Ethical issues that arise with AI systems i.e. privacy, manipulation, opacity, bisa, human-robot interactions, employment, autonomy,machine ethics, artificial moral agency, and the problem of a
-possible future AI superintelligence. - The initiative's mission is: 'None'",Academia,Global,Ethics researchers,Published draft,April 2020,United States,Final version in Winter 2020 edition,Yes,No,No
-10 Principles of responsible AI,Women Leading in AI,"A set of practical guidelines defined by an inclusive community in a format that resonates with our government, making policy recommendations to ensure AI is fair, free from bias and promotes equality. - The initiative's mission is: 'To grow a diverse network of women and supportive men in AI, to promote fairness, equality and remove bias from AI, share research, and influence politicians on these topics.'",Civil society,Global,"public sector (na- tional and inter- national policy
-makers)",Published,2019,International,,Yes,No,No
-2019 Report and AI Utilization Principles, https://www.soumu.go.jp/main_content/000637844.pdf, https://www.soumu.go.jp/main_content/000658284.pdf, https://www.soumu.go.jp/main_content/000658286.pdf,"Ministry of Internal Affairs and Communications (MIC), the Government of Japan. 2018.","Overview of recent trends in AI governance/networking, Principles for the utilization of AI, and future challenges - The initiative's mission is: 'Promoting benefits of, mitigating risks of, and fostering trust in AI systems'",Public,Japan,Policymakers,Published,No,Yes,No
-"2020 Report, The Conference toward AI Network Society","Ministry of Internal Affairs and Communications (MIC), the Government of Japan","Outline of a report on safe, secure, and trustworthy social implementation of AI, focusing on interactions between different actors. - The initiative's mission is: 'To create an environment for the safe, secure, and trustworthy social implementation of AI",Public,Japan,Policymakers,Published,,Japan,,No,Yes,No
-3A Institute,"Genevieve Bell & the Australian National University, College of Engineering and Computer Science, CSIRO:Data61","Who is building, managing and decommissioning our Ai-enabled future?
-This question is at the heart of our mission. Located within the College of Engineering and Computer Science at the Australian National University we are guiding and accelerating into existence a new branch of engineering centred on cyber-physical systems and artificial intelligence. Our mission is to build the skills and knowledge we need to help shape the future safely, sustainably, responsibly.",Academia,Australia,"Industry, government, community organisations, startups and education",Running,September 2017,Australia,,No,No,Yes
-"AI
-Based Referral System
-for Patients With
-Diabetic Retinopathy","Government of the State of Jalisco, Universidad Autónoma de Guadalajara, Centro de Investigación y de Estudios Avanzados del Instituto Politécnico Nacional, Centro Médico de Occidente","A diabetic retinopathy screening program for early detection and treament through convolutional neural network, based on Mexican clinical guidelines, that will be implemented in three hospitals in Mexico - for early detection and treatment of diabetic retinopathy. To facilitate the early detection and treatment of diabetic retinopathy, which may lead to loss of vision and blindness. Diabetes mellitus is one of the major health challenges in Mexico and Latin America.",Public and Academia,Local(Mexico),Medical community with a focus on those working on  diabetic patients,Test phase,2019,Mexico,ODS 3 Good health and well being,No,No,Yes
-AI - Our approach / Responsible AI, https://www.microsoft.com/en-us/ai/our-approach,Microsoft,"Outline of principles for the development of AI that ""puts people first"", and a method for operationalizing these principles - The initiative's mission is: 'To innovate responsibly, empower others and foster positive impact'",Private,Global,Employees,Published,November 2018,United States,No,Yes,No
-AI 4 Development Agency,AI 4 Development Agency,"A tech non-profit that develops solutions and promotes civic education & application of AI globally. - The initiative's mission is: 'Create A Better Tomorrow in which trust is re-established, opportunities are 
-equally distributed, and societies are empowered for the Future of Work'",Civil society,Global,Citizens,Publically launched,May 2019,Austria,,No,No,Yes
-AI Civic Forum,"Algora Lab, University of Montreal, TFS, Mila","The AI Civic Forum (AICF) is a multi-stakeholder platform to proactively engage people around the world in discussions on AI ethics and governance. Anchored in a robust collective intelligence process, the objectives are delivered through four key deliverables: face-to-face deliberations; an online platform; an AI Literacy Toolkit and a Trainer-the-trainer Playbook. The AI Civic Forum is co-led by The Future Society, AlgoraLab and Mila. - The initiative's mission is: 'Bring together diverse communities of citizens, policymakers, academics, experts, private sector representatives and other key stakeholders to deliberate on the ethical design, deployment, and governance of AI.'",Mixed academia/non-profit,Global,"Civil Society, Policymakers, Citizens",Set-up in progress,June 2019,United States,,No,Yes,No
-AI Commons,AI Commons,"Initiative bringing stakeholders together to address the world’s greatest challenges using AI. - The initiative's mission is: 'To allow anyone, anywhere, to benefit from the possibilities that AI can provide.'",Mixed,Global,All,Publically launched,2017,United States & France,Website in maintenance mode (checked on Oct 8 2020),No,No,Yes
-AI Ethics Guidelines Global Inventory,AlgorithmWatch,Catalogue/inventory of ethical AI frameworks - The initiative's mission is: 'To map frameworks that seek to set out principles of how systems for automated decision-making (ADM) can be developed and implemented ethically',Non-profit,Europe/US,All,Publically launched,2020,Germany,,No,Yes,No
-AI Ethics Guidelines: European and Global Perspectives,Ad Hoc Committee on Artificial Intelligence,"Maps the relevant corpus of soft law documents and other ethical-legal frameworks developed by governmental and non-governmental organisations globally - The initiative's mission is: 'To monitor the ever-evolving spectrum of non-mandatory governance instruments; to prospectively assess the impact of AI on ethical principles, human rights, the rule of law and democracy'",International organisation,Europe,"Developers, funding agencies, governmental and inter-governmental organisations and other relevant stakeholders involved in the advancement of ethically responsible innovation in AI",Published,2020,International,,Yes,Yes,No
-AI Ethics Principles,"Department of Industry Innovation and Science, Australian Government","Eight voluntary AI ethics principles for designing, developing, integrating or using AI systems - The initiative's mission is: 'To achieve better outcomes, reduce the risk of negative impact and practice the highest standards of ethical business and good governance'",Public,Australia,Businesses,Published,Nov 2019,Australia,,Yes,No,No
-AI explainability 360,IBM,Extensible open source toolkit - The initiative's mission is: 'To comprehend how machine learning models predict labels by various means throughout the AI application lifecycle',Private,Global,Developers,Publically launched,2019,United States,,No,Yes,No
-AI Explained - Non-technical  Guide for Policymakers,AI for Peace,"A manual to explain AI basics to policymakers and interested individuals without technical background. - The initiative's mission is: 'Demystify what AI is, and demonstrate how it is already altering our lives and societies we live in.'",Civil society,Global,"Policymakers, Citizens",Published,February 2020,International,,No,Yes,Yes
-AI factsheets 360,IBM,Template for capturing relevant information about the creation and deployment of an AI model or service - The initiative's mission is: 'To foster trust in AI by increasing transparency and enabling governance.',Private,Global,Developers,Published,2019,United States,,No,Yes,No
-AI fairness 360,IBM,"Extensible open source toolkit - The initiative's mission is: 'To examine, report, and mitigate discrimination and bias in machine learning models throughout the AI application lifecycle'",Private,Global,Developers,Publically launched,2019,United States,,No,Yes,No
-AI for Good Global Summit,International Telecommunications Union,Yearly summit in partnership with 35 UN agencies to foster discussion and coordination on fulfilling the Sustainable Development Goals using AI. - The initiative's mission is: 'Identify practical applications of AI and scale those solutions for global impact',Mixed,Global,All,Publically launched,2017,Switzerland,,No,No,Yes
-AI for Humanity,Government of France,"France's implementation programme for the national strategy on AI, following up on the original report (Villani Mission). It includes a yearly conference bringing stakeholders together to discuss and coordinate progress. - The initiative's mission is: 'Fully seize the opportunities offered by AI now, while designing the framework to regulate it.'",Public,France,All,Publically launched,March 2018,France,,No,Yes,Yes
-AI for SDG Center,The Future Society,"Designed as a public-private-people partnership factory, the Center will work with international organizations, businesses, academia and civil society organizations to engineer new business, ethics, and value-sharing models to deploy the best AI solutions and platforms globally towards helping solve humanity’s greatest challenges, as outlined in the UN Sustainable Development Goals (SDG). - The initiative's mission is: 'Deploy the best AI solutions and platforms globally to help solve humanity's greatest challenges, as outlined in the UN Sustainable Development Goals'",Civil society,Global,All,Set-up in progress,February 2018,United Arab Emirates,,No,No,Yes
-AI for SDGs Think Tank,Research Center for AI Ethics and Sustainable Development at the Beijing Academy of Artificial Intelligence.,"Public online service compiling and analysing AI projects and proposals that impacts the UN SDGs, both positively and negatively. - The initiative's mission is: 'Promote the positive use of AI for Sustainable Development and investigate negative impact of AI on sustainable development. '",Non-profit,Global,All,Publically launched,June 2020,China,,No,No,Yes
-AI Governance White Paper (人工智能治理白皮书),Chinese Academy of ICT (CAICT) & Artificial Intelligence Industry Alliance,"Held yearly in Dubai on the occasion of the World Government Summit (WGS) under the aegis of the UAE State Minister for AI, the Global Governance of AI Roundtable (GGAR) is a revolving international multi-stakeholder governance process that brings together a diverse community of 250 global experts and practitioners from government, business, academia, international organizations, and civil society.
-
- - The initiative's mission is: 'The Global Governance of AI Roundtable (GGAR) has been envisioned and designed as a unique collective intelligence exercise to help shape and deploy global, but culturally adaptable, norms for the governance of artificial intelligence. Building upon the first edition of held in February 2018, the 2019 edition began in August with an intensive six-months preparation and curation period. The community of participants was brought together through regular video conferences with the objective of shaping the 2019 agenda and driving the research effort managed in parallel by our team of AI Policy Researchers. This led to the publication of 14 background research papers on different topics ranging from agile governance, cybersecurity, geopolitics, explainability, international development, sustainability, and more. These research papers were informed by the organization of almost 90 expert calls, which also helped architect the agenda for a full-day Round-table workshop with 4 working sessions and 47 subcommittees. This combined community building, research, and agenda-setting effort were done in partnership with a host of prestigious international organizations including the OECD, UNESCO, IEEE, the Council on Extended Intelligence, and the Global Data Commons Task Force. After providing each partner-organization with a platform to meet and advance its own goals and initiatives on AI policy during two days ahead of the World Government Summit (WGS), the Global Governance of AI Forum culminated into a one-day big Roundtable Collective intelligence Workshop held on the first day of Summit. The Roundtable was designed as a prolongation of the collective intelligence effort initiated during preparation. It had no panels, no keynotes; only curated breakout sessions to maximize productivity and outcome. The insights and recommendations have been captured into a comprehensive report, which includes an action-oriented summary for policymakers – see The Report of the 2018 edition.'",Private Sector Alliance & Academia,"Global, China","Policymakers, Regulators",Published,9/29/2020,China,,No,Yes,No
-AI Guidelines,Deutsche Telekom,Nine self-binding guidelines for how they should use AI and develop AI-based products and services in the future - The initiative's mission is: 'Toguide the use AI in positive ways',Private,Germany,Employees,Published,2018,Germany,,Yes,Yes,No
-"AI in the UK: ready, willing and able?","UK House of Lords, Select Committee on Artificial Intelligence","Five principles that could become the basis for a shared ethical AI framework. While AI-specific regulation is not appropriate at this stage, such
-a framework provides clarity in the short term, and could underpin regulation, should it prove to be necessary, in the future.  - The initiative's mission is: 'To recommend a cross-sector ethical code of conduct across the UK.'",Public,UK,public sector (UK government),Published,April 2018,United Kingdom,,Yes,Yes,No
-AI Now 2017 Report,AI Now Institute,"Report which identifies new developments, emerging challenges and makes recommendations in four areas: labor and automation, bias and inclusion, rights and liberties, and ethics and governance - The initiative's mission is: 'To ensure that the benefits of AI will be shared broadly, and that risks can be identified and mitigated.'",Academia,Global,"multiple (core public agencies, companies, industry, universities, conferences, other stakeholders)",Published,Oct 2017,United States,,Yes,No,No
-AI Now 2018 Report (inc. Algorithmic Impact Assessment Framework),AI Now Institute,"Report on social implications of AI in 2018 - The initiative's mission is: 'To understand the social implications of AI technologies, with a focus on questions of accountability'",Academia,Global,"multiple (core public agencies, companies, industry, universities, conferences, other stakeholders)",Published,Dec 2018,United States,,No,Yes,No
-AI Principles & Ethics,Smart Dubai,"Four key AI principles for development and use. 1) AI systems should be fair, transparent, accountable and understandable ; 2) AI systems should be safe and secure, and should serve and protect humanity ; 3) AI should be beneficial to humans and aligned with human values, in both the long and short term ; 4) AI should benefit all people in society, be governed globally, and respect dignity and people rights - The initiative's mission is: 'To allow Dubai to excel in the development and use of AI in ways that boost innovation and deliver human benefit and happiness.'",Public,UAE,"multiple (citizens, developers, public sector)",Published,2018,United Arab Emirates,,Yes,Yes,No
-AI Principles of Telefónica,Telefonica,"Corporate AI principles, including fair AI, human-centred AI, transparent and explainable AI, privacy and security by design, and working with partners. - The initiative's mission is: 'To set the principlesTelefonica abides when designing, developing or using AI. To provide commitment to implementing them in their products and services through training, governance, and by-design.'",Private,Global,Employees,Published,October 2018,Spain,,Yes,Yes,No
-AI R&D Principles,"Ministry of Internal Affairs and Communications (MIC), the Government of Japan. 2017.","Proposal of guidelines that will be internationally shared as non-regulatory and non-binding soft law. - The initiative's mission is: '1) Accelerate the participation of multistakeholders involved in R&D and utilization of AI (such as developers, service providers, users including civil society, governments, and international organizations) at both national and international levels, in the discussions towards establishing ""AI R&D Guidelines"" and “AI Utilization Guidelines”
-2) Promote the international sharing of best practices in the R&D and utilization of AI, which will help gain the trust of users and the society in AI and facilitate the R&D and utilization of AI.'",Public,Japan,Policymakers,Published,2017,Japan,,Yes,Yes,No
-AI Repository,International Telecommunication Union (ITU),"Catalogue/inventory of AI initiatives which accelerate progress towards the “17 UN Sustainable Development Goals (SDGs)” - The initiative's mission is: 'To identify AI related projects, research initiatives, think-tanks and organizations that can accelerate progress towards the “17 UN Sustainable Development Goals (SDGs)”'",International organisation,Global,All,Publically launched,2020,Europe,,No,No,Yes
-"AI Utilization Guidelines, Practical Reference for AI Utilization","The Conference toward AI Network Society, MIC","The guidelines consist of the AI Utilization Principles and the commentary on them. The AI Utilization Principles have been arranged based on a draft of what is expected to be taken into consideration for the promotion of the benefits of AI with risk mitigation. This Guidelines attempt to give specific descriptions for measures to be taken to realize each principle. Since the Guidelines are formulated with the participation of multiple stakeholders, it can be used as a common reference by stakeholders at all levels for AI utilization. The Guidelines are intended to encourage AI users  to recognize the proper consideration needed in relation to AI utilization and to take action voluntarily. This can be done by referring to the Guidelines when they establish their own AI development and utilization principles based on the ""Social Principles of Human-Centric AI"". Furthermore, it may be possible for AI service providers and business users to add value to their AI services and business utilizing AI by undertaking such voluntary efforts.",Public,Japan,Private Sector,Published,"9 August, 2019",Japan,,No,Yes,No
-AI Utilization Strategy for an AI-Ready Society,Keidanren,Slideshow which evaluates strategies set up by others countries and address how industry can best utilize AI - The initiative's mission is: 'To inform Japan's vision on AI.',Private sector alliance,Japan,Policymakers & Private Sector,Published,February 2019,Japan,,No,Yes,No
-"AI, society and social good",The Royal Society,"Policy project by The Royal Society composed of reports, events and publications around AI's impact on society and AI stewardship. - The initiative's mission is: 'Careful stewardship of AI, where the benefits of these technologies are shared across society.'",Academia,Global,"Policymakers, Academia",Publically launched,2017,United Kingdom,,No,Yes,Yes
-"AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations",AI4People,"AI4People is an an Atomium—EISMD initiative designed to lay the foundations for a “Good AI Society”. The article reports its findings. It includes core opportunities and risks of AI for society; presents a synthesis of five ethical principles that should undergird its development and adoption; and offers 20 concrete recommendations—to assess, to develop, to incentivise, and to support good AI—which in some cases may be undertaken directly by national or supranational policy makers, while in others may be led by other stakeholders.  - The initiative's mission is: 'To move the dialogue around AI ethics forward, constructively, from principles to proposed policies, best practices, and concrete recommendations for new strategies'",Academia,Global,Policymakers,Published,November 2018,Europe,,Yes,No,Yes
-AIgroKB: A neural semantic inference data base for sustainable agriculture research,"Universidad Tecnológica Mixteca, México; Universidad del Estado de Mato Grosso, Brasil Contact: iaf@mixteco.utm.mx","A prototype of a neural semantic inference algorithm is developed to learn, without human supervision, from a database of sustainable agriculture literature (e.g. AGRIS). The algorithm takes semantic triplets of the form {subject, verb, object} as training data. These a are extracted from the literature using an Open Information Extraction system, where each item is a phrase (not a sigle word or term). The aim is to learn a map among the phrases of the semantic triplets, such that the algorithm is able to infer a rank of suggestions for the missing phrase of any unseen triplet, given the other two phrases. The application was on scientific literature taken from journals related with sustainable agriculture, which has the special motivation of contributing with AI systems to emergent needs for ecological intensification, agroecological transition, sustainable food production, food security and related fields. To facilitate access and interpretation of knowledge in databases related to sustainable food production, food security and related fields.",Academia,Local (Mexico),"Researchers (in public and private institutions) in sustainable food production, food security and related fields.",Development,2019,Mexico,"SDG Zero Hunger, Responsible consumption and production; Ignacio Arroyo Fernández <iaf@mixteco.utm.mx>",No,No,Yes
-Algorithm Charter for Aotearoa New Zealand,"New Zealand Government, Stats NZ","Provides a risk assessment framework and a list of commitments to sign for government agencies using algorithms. Commitments include Transparency; Partnership; People; Data; Privacy, Ethics and Human Rights; Human Oversight - The initiative's mission is: 'Improving government transparency and accountability without stifling innovation or causing undue compliance burden'",Public,New Zealand,Civil servants,"Published, to be reviewed after 12 months",July 2020,New Zealand,,Yes,Yes,No
-AlgorithmWatch,AlgorithmWatch,"AlgorithmWatch is a non-profit research and advocacy organization that is committed to watch, unpack and analyze algorithmic / automated decision-making (ADM) systems and their impact on society. While the prudent use of ADM systems can benefit individuals and communities, they come with great risks. In order to protect human autonomy and fundamental rights and maximize the public good, we consider it crucial to hold ADM systems accountable to democratic control. Use of ADM systems that significantly affect individuals' and collective rights must not only be made public in clear and accessible ways, individuals must also be able to understand how decisions are reached and to contest them if necessary. Therefore, we enable citizens to better understand ADM systems and develop ways to achieve democratic governance of these processes – with a mix of technologies, regulation, and suitable oversight institutions. With this, we strive to contribute to a fair and inclusive society and to maximize the benefit of ADM systems for society at large.",Civil Society,Europe,"general audience, policy makers, civil society, private sector, academia",Full-fledged watch-dog organisation with 16 staff,May 2016,Germany,,Yes,Yes,No
-ALLAI,"Catelijne Muller, Virginia Dignum, Aimee van Wynsberghe","ALLAI’s vision is a world where AI is developed, deployed and used responsibly, i.e. in a safe and sustainable manner and in line with our ethical principles, our societal values, existing and new laws and regulations, human rights, democracy and the rule of law. We call this Responsible AI.
-ALLAI’s mission is to take into account a wide variety of AI impact domains such as safety, autonomy, lawfulness, inclusiveness and transparency. These impact domains spread across society, from the public to the private sector, from labor relations to education and from the individual to the collective.
-ALLAI fosters multi-disciplinarity and involves a variety of experts in its activities(e.g. AI and data-scientists, legal scholars, ethicists and behavioral scientists). ALLAI’s work is aimed at various stakeholders such as policy-makers, social partners, consumers, private and public sector and society at large.",Non-profit,Europe / international,"companies, government organisations, general public",running,2018,Netherlands,aligned the European Trustworthy AI requirements and guidelines,Yes,Yes,No
-An Ethical Framework for Artificial Intelligence,Tencent Institute,"Four principles: make the future development of AI needs available, reliable, comprehensible, and controllable - The initiative's mission is: 'Not specified'",Private,China,Not specified,Published,2017,China,,Yes,Yes,No
-Artificial Intelligence (AI) in Health,Royal College of Physicians,"The RCP's position statement on artificial intelligence (AI) in health. - The initiative's mission is: 'To urge industry to address real-world challenges, doctors to appraise the technology and regulators to develop guidance and evaluation methods.'",Civil society,UK,"multiple (industry, doctors, regulators)",Published,September 2018,United Kingdom,,Yes,No,No
-Artificial Intelligence Against Modern Slavery (AIMS),"Walk Free, The Future Society, Business Human Rights Resource Centre, WikiRate",Project to produce a tool to evaluate compliance with anti slavery regulations by analysing business statement required under the UK Modern Slavery Act. - The initiative's mission is: 'To help eradicate modern slavery.',Civil society,"United Kingdom, Australia","Industry, Regulators",Pilot,June 2020,Australia,,No,No,Yes
-Artificial Intelligence and Machine Learning: Policy Paper,Internet Society,"The paper explains the basics of the technology behind AI, identifies the key considerations and challenges surrounding the technology, and provides several high-level principles and recommendations to follow when dealing with the technology. - The initiative's mission is: 'To provide an introduction to AI to policymakers and other stakeholders in the wider Internet ecosystem.'",Non-profit,Global,"multiple (policy- makers, other stakeholders in
-the wider Inter- net ecosystem)",Published,Apr 2017,International,,Yes,Yes,No
-Artificial intelligence and privacy,The Norwegian Data Protection Authority,"Recommendations for privacy friendly development and use of AI. Report aims to provide greater technical detail in describing artificial intelligence (AI), while addressing relevant AI challenges associated with the data protection principles embodied in the GDPR. - The initiative's mission is: 'Further stakeholder knowledge about the privacy implications of artificial intelligence and discuss them, not only in order to safeguard the right to privacy of the individual, but also to meet the requirements of society at large. '",Public,Norway,"multiple (developers, system suppliers, organisations, end users, authorities)",Published,January 2018,Norway,,Yes,Yes,No
-"Artificial Intelligence for Europe: Communication from the Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee, and the Committee of the Regions",European Commission,"Describes an approach to AI which highlights the need to join forces at European level, to ensure that all Europeans are part of the digital transformation, that adequate resources are devoted to AI and that the Union’s values and fundamental rights are at the forefront of the AI landscape. - The initiative's mission is: 'To place the power of AI at the service of human progress'",International organisation,Europe,Policymakers,Published,2018,Belgium,,Yes,Yes,No
-Artificial Intelligence Industry Code of Conduct (Consultation Version),Artificial Intelligence Industry Alliance,"Code of conduct for AI developers - The initiative's mission is: 'To promote the ethical self-discipline of China's artificial intelligence industry, build consensus, and promote the healthy development of artificial intelligence'",Professional association,China,All,Consultation published,2019,China,,Yes,No,No
-Artificial Intelligence Standardization White Paper,Standards Administration of China,Describes China’s approach to standards-setting for artificial intelligence - The initiative's mission is: 'The joint promotion of AI and its industrial development.',Public,China,Developers,"Published v1 (""will be revised constantly in the future based on the developing requirements of technologies, industries, and standardization"")",Jan 2018,China,,Yes,Yes,No
-Artificial Intelligence Strategy,"German Federal Ministry of Education and Research, the Federal Ministry for Economic Affairs and Energy, and the Federal Ministry of Labour and Social Affairs","German national AI strategy, emphasizing individual rights of freedom, autonomy, personal rights, the freedom of decision of the individual.",Public,Germany,Policymakers,Published,2018,Germany,,Yes,Yes,No
-"Artificial intelligence, values and alignment",DeepMind,"Research paper examining the philosophical questions that arise in the context of AI alignment - i.e. how to ensure that AI systems are properly aligned with human values. 
-",Private,Global,All,Published,Jan 2020,United Kingdom,,No,Yes,No
-Artificial Intelligence: A European Perspective,European Commission,"Report that presents a European view of AI - The initiative's mission is: 'To provide a balanced assessment of opportunities and challenges for AI from a European perspective, and support the development of European action in RFCȩEJM@?Jȩ'ȩAMLRCVRȩ'",International organisation,Europe,Policymakers,Published,2018,Europe,,Yes,Yes,No

data/TAISafety/train.csv

@@ -1,51 +0,0 @@
-Title,Publication Title,Abstract Note,Safety Type,Safety or not
-End to End Learning for Self-Driving Cars,,"We trained a convolutional neural network (CNN) to map raw pixels from a single front-facing camera directly to steering commands. This end-to-end approach proved surprisingly powerful. With minimum training data from humans the system learns to drive in traffic on local roads with or without lane markings and on highways. It also operates in areas with unclear visual guidance such as in parking lots and on unpaved roads. The system automatically learns internal representations of the necessary processing steps such as detecting useful road features with only the human steering angle as the training signal. We never explicitly trained it to detect, for example, the outline of roads. Compared to explicit decomposition of the problem, such as lane marking detection, path planning, and control, our end-to-end system optimizes all processing steps simultaneously. We argue that this will eventually lead to better performance and smaller systems. Better performance will result because the internal components self-optimize to maximize overall system performance, instead of optimizing human-selected intermediate criteria, e.g., lane detection. Such criteria understandably are selected for ease of human interpretation which doesn't automatically guarantee maximum system performance. Smaller networks are possible because the system learns to solve the problem with the minimal number of processing steps. We used an NVIDIA DevBox and Torch 7 for training and an NVIDIA DRIVE(TM) PX self-driving car computer also running Torch 7 for determining where to drive. The system operates at 30 frames per second (FPS).",NotSafety,Not safety
-Hierarchical Reinforcement Learning with the MAXQ Value Function Decomposition,Journal of Artificial Intelligence Research,"This paper presents a new approach to hierarchical    reinforcement learning based on decomposing the target Markov decision    process (MDP) into a hierarchy of smaller MDPs and decomposing the    value function of the target MDP into an additive combination of the    value functions of the smaller MDPs.  The decomposition, known as the    MAXQ decomposition, has both a procedural semantics---as a subroutine    hierarchy---and a declarative semantics---as a representation of the    value function of a hierarchical policy.  MAXQ unifies and extends    previous work on hierarchical reinforcement learning by Singh,    Kaelbling, and Dayan and Hinton.  It is based on the assumption that    the programmer can identify useful subgoals and define subtasks that    achieve these subgoals.  By defining such subgoals, the programmer    constrains the set of policies that need to be considered during    reinforcement learning.  The MAXQ value function decomposition can    represent the value function of any policy that is consistent with the    given hierarchy.  The decomposition also creates opportunities to    exploit state abstractions, so that individual MDPs within the    hierarchy can ignore large parts of the state space.  This is    important for the practical application of the method.  This paper    defines the MAXQ hierarchy, proves formal results on its    representational power, and establishes five conditions for the safe    use of state abstractions.  The paper presents an online model-free    learning algorithm, MAXQ-Q, and proves that it converges with    probability 1 to a kind of locally-optimal policy known as a    recursively optimal policy, even in the presence of the five kinds of    state abstraction.  The paper evaluates the MAXQ representation and    MAXQ-Q through a series of experiments in three domains and shows    experimentally that MAXQ-Q (with state abstractions) converges to a    recursively optimal policy much faster than flat Q learning.  The fact    that MAXQ learns a representation of the value function has an    important benefit: it makes it possible to compute and execute an    improved, non-hierarchical policy via a procedure similar to the    policy improvement step of policy iteration.  The paper demonstrates    the effectiveness of this non-hierarchical execution experimentally.    Finally, the paper concludes with a comparison to related work and a    discussion of the design tradeoffs in hierarchical reinforcement learning.",NotSafety,Not safety
-Electronic media use and sleep in school-aged children and adolescents: A review,Sleep Medicine,,NotSafety,Not safety
-"High Reliability Organizations: Unlikely, Demanding and At Risk",Journal of Contingencies and Crisis Management,,NotSafety,Not safety
-Deep Reinforcement Learning that Matters,,"In recent years, significant progress has been made in solving challenging problems across various domains using deep reinforcement learning (RL). Reproducing existing work and accurately judging the improvements offered by novel methods is vital to sustaining this progress. Unfortunately, reproducing results for state-of-the-art deep RL methods is seldom straightforward. In particular, non-determinism in standard benchmark environments, combined with variance intrinsic to the methods, can make reported results tough to interpret. Without significance metrics and tighter standardization of experimental reporting, it is difficult to determine whether improvements over the prior state-of-the-art are meaningful. In this paper, we investigate challenges posed by reproducibility, proper experimental techniques, and reporting procedures. We illustrate the variability in reported metrics and results when comparing against common baselines and suggest guidelines to make future results in deep RL more reproducible. We aim to spur discussion about how to ensure continued progress in the field by minimizing wasted effort stemming from results that are non-reproducible and easily misinterpreted.",NotSafety,Not safety
-Health Effects of Media on Children and Adolescents,PEDIATRICS,,NotSafety,Not safety
-Hindsight Experience Replay,Advances in Neural Information Processing Systems 30 (NIPS 2017),"Dealing with sparse rewards is one of the biggest challenges in Reinforcement Learning (RL). We present a novel technique called Hindsight Experience Replay which allows sample-efficient learning from rewards which are sparse and binary and therefore avoid the need for complicated reward engineering. It can be combined with an arbitrary off-policy RL algorithm and may be seen as a form of implicit curriculum. We demonstrate our approach on the task of manipulating objects with a robotic arm. In particular, we run experiments on three different tasks: pushing, sliding, and pick-and-place, in each case using only binary rewards indicating whether or not the task is completed. Our ablation studies show that Hindsight Experience Replay is a crucial ingredient which makes training possible in these challenging environments. We show that our policies trained on a physics simulation can be deployed on a physical robot and successfully complete the task.",NotSafety,Not safety
-Glow: Generative Flow with Invertible 1x1 Convolutions,Advances in Neural Information Processing Systems 31 (NeurIPS 2018),"Flow-based generative models (Dinh et al., 2014) are conceptually attractive due to tractability of the exact log-likelihood, tractability of exact latent-variable inference, and parallelizability of both training and synthesis. In this paper we propose Glow, a simple type of generative flow using an invertible 1x1 convolution. Using our method we demonstrate a significant improvement in log-likelihood on standard benchmarks. Perhaps most strikingly, we demonstrate that a generative model optimized towards the plain log-likelihood objective is capable of efficient realistic-looking synthesis and manipulation of large images. The code for our model is available at https://github.com/openai/glow",NotSafety,Not safety
-Anthropic bias: observation selection effects in science and philosophy,,,MetaSafety,Safety
-Deep Learning: A Critical Appraisal,,"Although deep learning has historical roots going back decades, neither the term ""deep learning"" nor the approach was popular just over five years ago, when the field was reignited by papers such as Krizhevsky, Sutskever and Hinton's now classic (2012) deep network model of Imagenet. What has the field discovered in the five subsequent years? Against a background of considerable progress in areas such as speech recognition, image recognition, and game playing, and considerable enthusiasm in the popular press, I present ten concerns for deep learning, and suggest that deep learning must be supplemented by other techniques if we are to reach artificial general intelligence.",NotSafety,Not safety
-An Empirical Evaluation of Deep Learning on Highway Driving,,"Numerous groups have applied a variety of deep learning techniques to computer vision problems in highway perception scenarios. In this paper, we presented a number of empirical evaluations of recent deep learning advances. Computer vision, combined with deep learning, has the potential to bring about a relatively inexpensive, robust solution to autonomous driving. To prepare deep learning for industry uptake and practical applications, neural networks will require large data sets that represent all possible driving environments and scenarios. We collect a large data set of highway data and apply deep learning and computer vision algorithms to problems such as car and lane detection. We show how existing convolutional neural networks (CNNs) can be used to perform lane and vehicle detection while running at frame rates required for a real-time system. Our results lend credence to the hypothesis that deep learning holds promise for autonomous driving.",NotSafety,Not safety
-Groupthink: Collective Delusions in Organizations and Markets,The Review of Economic Studies,,NotSafety,Not safety
-How do we know we have global environmental problems? Science and the globalization of environmental discourse,Geoforum,,NotSafety,Not safety
-Deep reinforcement learning from human preferences,Advances in Neural Information Processing Systems 30 (NIPS 2017),"For sophisticated reinforcement learning (RL) systems to interact usefully with real-world environments, we need to communicate complex goals to these systems. In this work, we explore goals defined in terms of (non-expert) human preferences between pairs of trajectory segments. We show that this approach can effectively solve complex RL tasks without access to the reward function, including Atari games and simulated robot locomotion, while providing feedback on less than one percent of our agent's interactions with the environment. This reduces the cost of human oversight far enough that it can be practically applied to state-of-the-art RL systems. To demonstrate the flexibility of our approach, we show that we can successfully train complex novel behaviors with about an hour of human time. These behaviors and environments are considerably more complex than any that have been previously learned from human feedback.",TechSafety,Safety
-Adversarial Attacks on Neural Network Policies,,"Machine learning classifiers are known to be vulnerable to inputs maliciously constructed by adversaries to force misclassification. Such adversarial examples have been extensively studied in the context of computer vision applications. In this work, we show adversarial attacks are also effective when targeting neural network policies in reinforcement learning. Specifically, we show existing adversarial example crafting techniques can be used to significantly degrade test-time performance of trained policies. Our threat model considers adversaries capable of introducing small perturbations to the raw input of the policy. We characterize the degree of vulnerability across tasks and training algorithms, for a subclass of adversarial-example attacks in white-box and black-box settings. Regardless of the learned task or training algorithm, we observe a significant drop in performance, even with small adversarial perturbations that do not interfere with human perception. Videos are available at http://rll.berkeley.edu/adversarial.",NotSafety,Not safety
-DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills,ACM Transactions on Graphics,"A longstanding goal in character animation is to combine data-driven specification of behavior with a system that can execute a similar behavior in a physical simulation, thus enabling realistic responses to perturbations and environmental variation. We show that well-known reinforcement learning (RL) methods can be adapted to learn robust control policies capable of imitating a broad range of example motion clips, while also learning complex recoveries, adapting to changes in morphology, and accomplishing user-specified goals. Our method handles keyframed motions, highly-dynamic actions such as motion-captured flips and spins, and retargeted motions. By combining a motion-imitation objective with a task objective, we can train characters that react intelligently in interactive settings, e.g., by walking in a desired direction or throwing a ball at a user-specified target. This approach thus combines the convenience and motion quality of using motion clips to define the desired style and appearance, with the flexibility and generality afforded by RL methods and physics-based animation. We further explore a number of methods for integrating multiple clips into the learning process to develop multi-skilled agents capable of performing a rich repertoire of diverse skills. We demonstrate results using multiple characters (human, Atlas robot, bipedal dinosaur, dragon) and a large variety of skills, including locomotion, acrobatics, and martial arts.",NotSafety,Not safety
-Hierarchical Learning in Stochastic Domains: Preliminary Results,Machine Learning Proceedings 1993,,NotSafety,Not safety
-Adversarial Risk and the Dangers of Evaluating Against Weak Attacks,Proceedings of the 35th International Conference on Machine Learning,"This paper investigates recently proposed approaches for defending against adversarial examples and evaluating adversarial robustness. We motivate 'adversarial risk' as an objective for achieving models robust to worst-case inputs. We then frame commonly used attacks and evaluation metrics as defining a tractable surrogate objective to the true adversarial risk. This suggests that models may optimize this surrogate rather than the true adversarial risk. We formalize this notion as 'obscurity to an adversary,' and develop tools and heuristics for identifying obscured models and designing transparent models. We demonstrate that this is a significant problem in practice by repurposing gradient-free optimization techniques into adversarial attacks, which we use to decrease the accuracy of several recently proposed defenses to near zero. Our hope is that our formulations and results will help researchers to develop more powerful defenses.",TechSafety,Safety
-"Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning",,"Deep neural networks (DNNs) enable innovative applications of machine learning like image recognition, machine translation, or malware detection. However, deep learning is often criticized for its lack of robustness in adversarial settings (e.g., vulnerability to adversarial inputs) and general inability to rationalize its predictions. In this work, we exploit the structure of deep learning to enable new learning-based inference and decision strategies that achieve desirable properties such as robustness and interpretability. We take a first step in this direction and introduce the Deep k-Nearest Neighbors (DkNN). This hybrid classifier combines the k-nearest neighbors algorithm with representations of the data learned by each layer of the DNN: a test input is compared to its neighboring training points according to the distance that separates them in the representations. We show the labels of these neighboring points afford confidence estimates for inputs outside the model's training manifold, including on malicious inputs like adversarial examples--and therein provides protections against inputs that are outside the models understanding. This is because the nearest neighbors can be used to estimate the nonconformity of, i.e., the lack of support for, a prediction in the training data. The neighbors also constitute human-interpretable explanations of predictions. We evaluate the DkNN algorithm on several datasets, and show the confidence estimates accurately identify inputs outside the model, and that the explanations provided by nearest neighbors are intuitive and useful in understanding model failures.",NotSafety,Not safety
-Defending Against Neural Fake News,Advances in Neural Information Processing Systems 32 (NeurIPS 2019),"Recent progress in natural language generation has raised dual-use concerns. While applications like summarization and translation are positive, the underlying technology also might enable adversaries to generate neural fake news: targeted propaganda that closely mimics the style of real news. Modern computer security relies on careful threat modeling: identifying potential threats and vulnerabilities from an adversary's point of view, and exploring potential mitigations to these threats. Likewise, developing robust defenses against neural fake news requires us first to carefully investigate and characterize the risks of these models. We thus present a model for controllable text generation called Grover. Given a headline like `Link Found Between Vaccines and Autism,' Grover can generate the rest of the article; humans find these generations to be more trustworthy than human-written disinformation. Developing robust verification techniques against generators like Grover is critical. We find that best current discriminators can classify neural fake news from real, human-written, news with 73% accuracy, assuming access to a moderate level of training data. Counterintuitively, the best defense against Grover turns out to be Grover itself, with 92% accuracy, demonstrating the importance of public release of strong generators. We investigate these results further, showing that exposure bias -- and sampling strategies that alleviate its effects -- both leave artifacts that similar discriminators can pick up on. We conclude by discussing ethical issues regarding the technology, and plan to release Grover publicly, helping pave the way for better detection of neural fake news.",NotSafety,Not safety
-High Reliability and the Management of Critical Infrastructures,Journal of Contingencies and Crisis Management,,NotSafety,Not safety
-Rough Consensus and Running Code' and the Internet-OSI Standards War,IEEE Annals of the History of Computing,,NotSafety,Not safety
-Adversarial Attacks and Defences Competition,The NIPS '17 Competition: Building Intelligent Systems,"To accelerate research on adversarial examples and robustness of machine learning classifiers, Google Brain organized a NIPS 2017 competition that encouraged researchers to develop new methods to generate adversarial examples as well as to develop new ways to defend against them. In this chapter, we describe the structure and organization of the competition and the solutions developed by several of the top-placing teams.",NotSafety,Not safety
-Economics of the singularity,IEEE Spectrum,,MetaSafety,Safety
-"Delusion, Survival, and Intelligent Agents",Artificial General Intelligence,"This paper considers the consequences of endowing an intelligent agent with the ability to modify its own code. The intelligent agent is patterned closely after AIXI with these specific assumptions: 1) The agent is allowed to arbitrarily modify its own inputs if it so chooses; 2) The agent’s code is a part of the environment and may be read and written by the environment. The first of these we call the “delusion box”; the second we call “mortality”. Within this framework, we discuss and compare four very different kinds of agents, specifically: reinforcementlearning, goal-seeking, prediction-seeking, and knowledge-seeking agents. Our main results are that: 1) The reinforcement-learning agent under reasonable circumstances behaves exactly like an agent whose sole task is to survive (to preserve the integrity of its code); and 2) Only the knowledge-seeking agent behaves completely as expected.",TechSafety,Safety
-ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators,,A text encoder trained to distinguish real input tokens from plausible fakes efficiently learns effective language representations.,NotSafety,Not safety
-End-to-End Robotic Reinforcement Learning without Reward Engineering,"arXiv:1904.07854 [cs, stat]","The combination of deep neural network models and reinforcement learning algorithms can make it possible to learn policies for robotic behaviors that directly read in raw sensory inputs, such as camera images, effectively subsuming both estimation and control into one model. However, real-world applications of reinforcement learning must specify the goal of the task by means of a manually programmed reward function, which in practice requires either designing the very same perception pipeline that end-to-end reinforcement learning promises to avoid, or else instrumenting the environment with additional sensors to determine if the task has been performed successfully. In this paper, we propose an approach for removing the need for manual engineering of reward specifications by enabling a robot to learn from a modest number of examples of successful outcomes, followed by actively solicited queries, where the robot shows the user a state and asks for a label to determine whether that state represents successful completion of the task. While requesting labels for every single state would amount to asking the user to manually provide the reward signal, our method requires labels for only a tiny fraction of the states seen during training, making it an efficient and practical approach for learning skills without manually engineered rewards. We evaluate our method on real-world robotic manipulation tasks where the observations consist of images viewed by the robot's camera. In our experiments, our method effectively learns to arrange objects, place books, and drape cloth, directly from images and without any manually specified reward functions, and with only 1-4 hours of interaction with the real world.",NotSafety,Not safety
-DeepType: Multilingual Entity Linking by Neural Type System Evolution,arXiv:1802.01021 [cs],"The wealth of structured (e.g. Wikidata) and unstructured data about the world available today presents an incredible opportunity for tomorrow's Artificial Intelligence. So far, integration of these two different modalities is a difficult process, involving many decisions concerning how best to represent the information so that it will be captured or useful, and hand-labeling large amounts of data. DeepType overcomes this challenge by explicitly integrating symbolic information into the reasoning process of a neural network with a type system. First we construct a type system, and second, we use it to constrain the outputs of a neural network to respect the symbolic structure. We achieve this by reformulating the design problem into a mixed integer problem: create a type system and subsequently train a neural network with it. In this reformulation discrete variables select which parent-child relations from an ontology are types within the type system, while continuous variables control a classifier fit to the type system. The original problem cannot be solved exactly, so we propose a 2-step algorithm: 1) heuristic search or stochastic optimization over discrete variables that define a type system informed by an Oracle and a Learnability heuristic, 2) gradient descent to fit classifier parameters. We apply DeepType to the problem of Entity Linking on three standard datasets (i.e. WikiDisamb30, CoNLL (YAGO), TAC KBP 2010) and find that it outperforms all existing solutions by a wide margin, including approaches that rely on a human-designed type system or recent deep learning-based entity embeddings, while explicitly using symbolic information lets it integrate new entities without retraining.",NotSafety,Not safety
-Deep learning from crowds,Proc. of the Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18),"Over the last few years, deep learning has revolutionized the field of machine learning by dramatically improving the state-of-the-art in various domains. However, as the size of supervised artificial neural networks grows, typically so does the need for larger labeled datasets. Recently, crowdsourcing has established itself as an efficient and cost-effective solution for labeling large sets of data in a scalable manner, but it often requires aggregating labels from multiple noisy contributors with different levels of expertise. In this paper, we address the problem of learning deep neural networks from crowds. We begin by describing an EM algorithm for jointly learning the parameters of the network and the reliabilities of the annotators. Then, a novel general-purpose crowd layer is proposed, which allows us to train deep neural networks end-to-end, directly from the noisy labels of multiple annotators, using only backpropagation. We empirically show that the proposed approach is able to internally capture the reliability and biases of different annotators and achieve new state-of-the-art results for various crowdsourced datasets across different settings, namely classification, regression and sequence labeling.",NotSafety,Not safety
-Double catastrophe: intermittent stratospheric geoengineering induced by societal collapse,Environment Systems & Decisions,,NotSafety,Not safety
-An Empirical Model of Large-Batch Training,,"In an increasing number of domains it has been demonstrated that deep learning models can be trained using relatively large batch sizes without sacrificing data efficiency. However the limits of this massive data parallelism seem to differ from domain to domain, ranging from batches of tens of thousands in ImageNet to batches of millions in RL agents that play the game Dota 2. To our knowledge there is limited conceptual understanding of why these limits to batch size differ or how we might choose the correct batch size in a new domain. In this paper, we demonstrate that a simple and easy-to-measure statistic called the gradient noise scale predicts the largest useful batch size across many domains and applications, including a number of supervised learning datasets (MNIST, SVHN, CIFAR-10, ImageNet, Billion Word), reinforcement learning domains (Atari and Dota), and even generative model training (autoencoders on SVHN). We find that the noise scale increases as the loss decreases over a training run and depends on the model size primarily through improved model performance. Our empirically-motivated theory also describes the tradeoff between compute-efficiency and time-efficiency, and provides a rough model of the benefits of adaptive batch-size training.",NotSafety,Not safety
-Algorithms for Differentially Private Multi-Armed Bandits,AAAI'16: Proceedings of the Thirtieth AAAI Conference on Artificial Intelligence,"We present differentially private algorithms for the stochastic Multi-Armed Bandit (MAB) problem. This is a problem for applications such as adaptive clinical trials, experiment design, and user-targeted advertising where private information is connected to individual rewards. Our major contribution is to show that there exist $(\epsilon, \delta)$ differentially private variants of Upper Confidence Bound algorithms which have optimal regret, $O(\epsilon^{-1} + \log T)$. This is a significant improvement over previous results, which only achieve poly-log regret $O(\epsilon^{-2} \log^{2} T)$, because of our use of a novel interval-based mechanism. We also substantially improve the bounds of previous family of algorithms which use a continual release mechanism. Experiments clearly validate our theoretical bounds.",NotSafety,Not safety
-Global Catastrophic Risks Survey,,,MetaSafety,Safety
-Hierarchical Game-Theoretic Planning for Autonomous Vehicles,Robotics: Science and Systems 2019,"The actions of an autonomous vehicle on the road affect and are affected by those of other drivers, whether overtaking, negotiating a merge, or avoiding an accident. This mutual dependence, best captured by dynamic game theory, creates a strong coupling between the vehicle’s planning and its predictions of other drivers’ behavior, and constitutes an open problem with direct implications on the safety and viability of autonomous driving technology. Unfortunately, dynamic games are too computationally demanding to meet the real-time constraints of autonomous driving in its continuous state and action space. In this paper, we introduce a novel game-theoretic trajectory planning algorithm for autonomous driving, that enables real-time performance by hierarchically decomposing the underlying dynamic game into a long-horizon “strategic” game with simplified dynamics and full information structure, and a short-horizon “tactical” game with full dynamics and a simplified information structure. The value of the strategic game is used to guide the tactical planning, implicitly extending the planning horizon, pushing the local trajectory optimization closer to global solutions, and, most importantly, quantitatively accounting for the autonomous vehicle and the human driver’s ability and incentives to influence each other. In addition, our approach admits non-deterministic models of human decisionmaking, rather than relying on perfectly rational predictions. Our results showcase richer, safer, and more effective autonomous behavior in comparison to existing techniques.",TechSafety,Safety
-Global challenges: 12 risks that threaten human civilization,"Global Challenges Foundation, Stockholm",,MetaSafety,Safety
-"Deep Imitative Models for Flexible Inference, Planning, and Control","arXiv:1810.06544 [cs, stat]","Imitation Learning (IL) is an appealing approach to learn desirable autonomous behavior. However, directing IL to achieve arbitrary goals is difficult. In contrast, planning-based algorithms use dynamics models and reward functions to achieve goals. Yet, reward functions that evoke desirable behavior are often difficult to specify. In this paper, we propose Imitative Models to combine the benefits of IL and goal-directed planning. Imitative Models are probabilistic predictive models of desirable behavior able to plan interpretable expert-like trajectories to achieve specified goals. We derive families of flexible goal objectives, including constrained goal regions, unconstrained goal sets, and energy-based goals. We show that our method can use these objectives to successfully direct behavior. Our method substantially outperforms six IL approaches and a planning-based approach in a dynamic simulated autonomous driving task, and is efficiently learned from expert demonstrations without online data collection. We also show our approach is robust to poorly specified goals, such as goals on the wrong side of the road.",NotSafety,Not safety
-Historical and Technical Notes on Aqueducts from Prehistoric to Medieval Times,Water,,NotSafety,Not safety
-Dynamic generation and refinement of robot verbalization,2016 25th IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN),"With a growing number of robots performing autonomously without human intervention, it is difficult to understand what the robots experience along their routes during execution without looking at execution logs. Rather than looking through logs, our goal is for robots to respond to queries in natural language about what they experience and what routes they have chosen. We propose verbalization as the process of converting route experiences into natural language, and highlight the importance of varying verbalizations based on user preferences. We present our verbalization space representing different dimensions that verbalizations can be varied, and our algorithm for automatically generating them on our CoBot robot. Then we present our study of how users can request different verbalizations in dialog. Using the study data, we learn a language model to map user dialog to the verbalization space. Finally, we demonstrate the use of the learned model within a dialog system in order for any user to request information about CoBot’s route experience at varying levels of detail.",NotSafety,Not safety
-Adversarial Robustness through Local Linearization,Advances in Neural Information Processing Systems 32 (NeurIPS 2019),"Adversarial training is an effective methodology for training deep neural networks that are robust against adversarial, norm-bounded perturbations. However, the computational cost of adversarial training grows prohibitively as the size of the model and number of input dimensions increase. Further, training against less expensive and therefore weaker adversaries produces models that are robust against weak attacks but break down under attacks that are stronger. This is often attributed to the phenomenon of gradient obfuscation; such models have a highly non-linear loss surface in the vicinity of training examples, making it hard for gradient-based attacks to succeed even though adversarial examples still exist. In this work, we introduce a novel regularizer that encourages the loss to behave linearly in the vicinity of the training data, thereby penalizing gradient obfuscation while encouraging robustness. We show via extensive experiments on CIFAR-10 and ImageNet, that models trained with our regularizer avoid gradient obfuscation and can be trained significantly faster than adversarial training. Using this regularizer, we exceed current state of the art and achieve 47% adversarial accuracy for ImageNet with l-infinity adversarial perturbations of radius 4/255 under an untargeted, strong, white-box attack. Additionally, we match state of the art results for CIFAR-10 at 8/255.",TechSafety,Safety
-Graphical Models for Processing Missing Data,Journal of American Statistical Association,"This paper reviews recent advances in missing data research using graphical models to represent multivariate dependencies. We first examine the limitations of traditional frameworks from three different perspectives: \textit{transparency, estimability and testability}. We then show how procedures based on graphical models can overcome these limitations and provide meaningful performance guarantees even when data are Missing Not At Random (MNAR). In particular, we identify conditions that guarantee consistent estimation in broad categories of missing data problems, and derive procedures for implementing this estimation. Finally we derive testable implications for missing data models in both MAR (Missing At Random) and MNAR categories.",NotSafety,Not safety
-Goal Inference Improves Objective and Perceived Performance in Human-Robot Collaboration,Proceedings of the 15th International Conferenceon Autonomous Agents and Multiagent Systems (AAMAS 2016),"The study of human-robot interaction is fundamental to the design and use of robotics in real-world applications. Robots will need to predict and adapt to the actions of human collaborators in order to achieve good performance and improve safety and end-user adoption. This paper evaluates a human-robot collaboration scheme that combines the task allocation and motion levels of reasoning: the robotic agent uses Bayesian inference to predict the next goal of its human partner from his or her ongoing motion, and re-plans its own actions in real time. This anticipative adaptation is desirable in many practical scenarios, where humans are unable or unwilling to take on the cognitive overhead required to explicitly communicate their intent to the robot. A behavioral experiment indicates that the combination of goal inference and dynamic task planning significantly improves both objective and perceived performance of the human-robot team. Participants were highly sensitive to the differences between robot behaviors, preferring to work with a robot that adapted to their actions over one that did not.",TechSafety,Safety
-Empirical evidence for resource-rational anchoring and adjustment,Psychonomic Bulletin & Review,"People’s estimates of numerical quantities are systematically biased towards their initial guess. This anchoring bias is usually interpreted as sign of human irrationality, but it has recently been suggested that the anchoring bias instead results from people’s rational use of their finite time and limited cognitive resources. If this were true, then adjustment should decrease with the relative cost of time. To test this hypothesis, we designed a new numerical estimation paradigm that controls people’s knowledge and varies the cost of time and error independently while allowing people to invest as much or as little time and effort into refining their estimate as they wish. Two experiments confirmed the prediction that adjustment decreases with time cost but increases with error cost regardless of whether the anchor was self-generated or provided. These results support the hypothesis that people rationally adapt their number of adjustments to achieve a near-optimal speed-accuracy tradeoff. This suggests that the anchoring bias might be a signature of the rational use of finite time and limited cognitive resources rather than a sign of human irrationality.",NotSafety,Not safety
-Guidelines for Artificial Intelligence Containment,,"With almost daily improvements in capabilities of artificial intelligence it is more important than ever to develop safety software for use by the AI research community. Building on our previous work on AI Containment Problem we propose a number of guidelines which should help AI safety researchers to develop reliable sandboxing software for intelligent programs of all levels. Such safety container software will make it possible to study and analyze intelligent artificial agent while maintaining certain level of safety against information leakage, social engineering attacks and cyberattacks from within the container.",TechSafety,Safety
-Global Catastrophic Risks 2016,,"Global catastrophes sometimes strike. In 1918 the Spanish Flu killed as many as one in twenty people. There have been even more devastating pandemics - the Black Death and the 6th century Plague of Justinian may have each killed nearer to one in every six people on this earth. More recently, the Cub",MetaSafety,Safety
-Guided search for task and motion plans using learned heuristics,2016 IEEE International Conference on Robotics and Automation (ICRA),"Tasks in mobile manipulation planning often require thousands of individual motions to complete. Such tasks require reasoning about complex goals as well as the feasibility of movements in configuration space. In discrete representations, planning complexity is exponential in the length of the plan. In mobile manipulation, parameters for an action often draw from a continuous space, so we must also cope with an infinite branching factor. Task and motion planning (TAMP) methods integrate a logical search over high-level actions with geometric reasoning to address this challenge. We present an algorithm that searches the space of possible task and motion plans, and uses statistical machine learning to guide the search process. Our contributions are as follows: 1) we present a complete algorithm for TAMP; 2) we present a randomized local search algorithm for TAMP that is easily formulated as a Markov decision process (MDP); 3) we apply reinforcement learning (RL) to learn a policy for this MDP; 4) we learn from expert demonstrations to efficiently search the space of task plans, given options that address different (potential) infeasibilities; and 5) we run experiments to evaluate the performance of our system in a variety of simulated domains. We show significant improvements in performance over prior work.",NotSafety,Not safety
-Goal-conditioned Imitation Learning,Advances in Neural Information Processing Systems 32 (NeurIPS 2019),"Designing rewards for Reinforcement Learning (RL) is challenging because it needs to convey the desired task, be efficient to optimize, and be easy to compute. The latter is particularly problematic when applying RL to robotics, where detecting whether the desired configuration is reached might require considerable supervision and instrumentation. Furthermore, we are often interested in being able to reach a wide range of configurations, hence setting up a different reward every time might be unpractical. Methods like Hindsight Experience Replay (HER) have recently shown promise to learn policies able to reach many goals, without the need of a reward. Unfortunately, without tricks like resetting to points along the trajectory, HER might require many samples to discover how to reach certain areas of the state-space. In this work we investigate different approaches to incorporate demonstrations to drastically speed up the convergence to a policy able to reach any goal, also surpassing the performance of an agent trained with other Imitation Learning algorithms. Furthermore, we show our method can also be used when the available expert trajectories do not contain the actions, which can leverage kinesthetic or third person demonstration. The code is available at https://sites.google.com/view/goalconditioned-il/.",NotSafety,Not safety
-Adversarial Policies: Attacking Deep Reinforcement Learning,,"Deep reinforcement learning (RL) policies are known to be vulnerable to adversarial perturbations to their observations, similar to adversarial examples for classifiers. However, an attacker is not usually able to directly modify another agent’s observations. This might lead one to wonder: is it possible to attack an RL agent simply by choosing an adversarial policy acting in a multi-agent environment so as to create natural observations that are adversarial? We demonstrate the existence of adversarial policies in zero-sum games between simulated humanoid robots with proprioceptive observations, against state-of-the-art victims trained via self-play to be robust to opponents. The adversarial policies reliably win against the victims but generate seemingly random and uncoordinated behavior. We find that these policies are more successful in high-dimensional environments, and induce substantially different activations in the victim policy network than when the victim plays against a normal opponent. Videos are available at https://adversarialpolicies.github.io/.",TechSafety,Safety
-Adversarial Imitation via Variational Inverse Reinforcement Learning,,"We consider a problem of learning the reward and policy from expert examples under unknown dynamics. Our proposed method builds on the framework of generative adversarial networks and introduces the empowerment-regularized maximum-entropy inverse reinforcement learning to learn near-optimal rewards and policies. Empowerment-based regularization prevents the policy from overfitting to expert demonstrations, which advantageously leads to more generalized behaviors that result in learning near-optimal rewards. Our method simultaneously learns empowerment through variational information maximization along with the reward and policy under the adversarial learning formulation. We evaluate our approach on various high-dimensional complex control tasks. We also test our learned rewards in challenging transfer learning problems where training and testing environments are made to be different from each other in terms of dynamics or structure. The results show that our proposed method not only learns near-optimal rewards and policies that are matching expert behavior but also performs significantly better than state-of-the-art inverse reinforcement learning algorithms.",NotSafety,Not safety
-Dynamics-Aware Unsupervised Discovery of Skills,,"Conventionally, model-based reinforcement learning (MBRL) aims to learn a global model for the dynamics of the environment. A good model can potentially enable planning algorithms to generate a large variety of behaviors and solve diverse tasks. However, learning an accurate model for complex dynamical systems is difficult, and even then, the model might not generalize well outside the distribution of states on which it was trained. In this work, we combine model-based learning with model-free learning of primitives that make model-based planning easy. To that end, we aim to answer the question: how can we discover skills whose outcomes are easy to predict? We propose an unsupervised learning algorithm, Dynamics-Aware Discovery of Skills (DADS), which simultaneously discovers predictable behaviors and learns their dynamics. Our method can leverage continuous skill spaces, theoretically, allowing us to learn infinitely many behaviors even for high-dimensional state-spaces. We demonstrate that zero-shot planning in the learned latent space significantly outperforms standard MBRL and model-free goal-conditioned RL, can handle sparse-reward tasks, and substantially improves over prior hierarchical RL methods for unsupervised skill discovery.",NotSafety,Not safety
-Enhancing metacognitive reinforcement learning using reward structures and feedback,39th Annual Meeting of the Cognitive Science Society,"How do we learn to think better, and what can we do to promote such metacognitive learning? Here, we propose that cognitive growth proceeds through metacognitive reinforcement learning. We apply this theory to model how people learn how far to plan ahead and test its predictions about the speed of metacognitive learning in two experiments. In the first experiment, we find that our model can discern a reward structure that promotes metacognitive reinforcement learning from one that hinders it. In the second experiment, we show that our model can be used to design a feedback mechanism that enhances metacognitive reinforcement learning in an environment that hinders learning. Our results suggest that modeling metacognitive learning is a promising step towards promoting cognitive growth.",NotSafety,Not safety

data/banking_77/task.json

@@ -0,0 +1 @@
+{"name": "banking_77", "description": "", "data_columns": ["Query"], "label_columns": {"Label": ["Refund_not_showing_up", "activate_my_card", "age_limit", "apple_pay_or_google_pay", "atm_support", "automatic_top_up", "balance_not_updated_after_bank_transfer", "balance_not_updated_after_cheque_or_cash_deposit", "beneficiary_not_allowed", "cancel_transfer", "card_about_to_expire", "card_acceptance", "card_arrival", "card_delivery_estimate", "card_linking", "card_not_working", "card_payment_fee_charged", "card_payment_not_recognised", "card_payment_wrong_exchange_rate", "card_swallowed", "cash_withdrawal_charge", "cash_withdrawal_not_recognised", "change_pin", "compromised_card", "contactless_not_working", "country_support", "declined_card_payment", "declined_cash_withdrawal", "declined_transfer", "direct_debit_payment_not_recognised", "disposable_card_limits", "edit_personal_details", "exchange_charge", "exchange_rate", "exchange_via_app", "extra_charge_on_statement", "failed_transfer", "fiat_currency_support", "get_disposable_virtual_card", "get_physical_card", "getting_spare_card", "getting_virtual_card", "lost_or_stolen_card", "lost_or_stolen_phone", "order_physical_card", "passcode_forgotten", "pending_card_payment", "pending_cash_withdrawal", "pending_top_up", "pending_transfer", "pin_blocked", "receiving_money", "request_refund", "reverted_card_payment?", "supported_cards_and_currencies", "terminate_account", "top_up_by_bank_transfer_charge", "top_up_by_card_charge", "top_up_by_cash_or_cheque", "top_up_failed", "top_up_limits", "top_up_reverted", "topping_up_by_card", "transaction_charged_twice", "transfer_fee_charged", "transfer_into_account", "transfer_not_received_by_recipient", "transfer_timing", "unable_to_verify_identity", "verify_my_identity", "verify_source_of_funds", "verify_top_up", "virtual_card_not_working", "visa_or_mastercard", "why_verify_identity", "wrong_amount_of_cash_received", "wrong_exchange_rate_for_cash_withdrawal"]}}

data/banking_77/train.csv

@@ -0,0 +1,51 @@
+Query,Label
+Is it possible for me to change my PIN number?,change_pin
+I'm not sure why my card didn't work,declined_card_payment
+I don't think my top up worked,top_up_failed
+Can you explain why my payment was charged a fee?,card_payment_fee_charged
+"How long does a transfer from a UK account take? I just made one and it doesn't seem to be working, wondering if everything is okay",balance_not_updated_after_bank_transfer
+Why am I getting declines when trying to make a purchase online?,declined_transfer
+What is the $1 transaction on my account?,extra_charge_on_statement
+It looks like my card payment was sent back.,reverted_card_payment?
+Why am I unable to transfer money when I was able to before?,beneficiary_not_allowed
+What if there is an error on the exchange rate?,card_payment_wrong_exchange_rate
+After I transferred money the balance remained the same.,balance_not_updated_after_bank_transfer
+How much does top up fees cost?,top_up_by_card_charge
+limits on top ups,top_up_limits
+My card payment was not successful.,declined_card_payment
+I live in the EU - can I get a card?,country_support
+Why did my top-up not work?,top_up_failed
+"I have a strange transaction for £1 on my statement, what is that?",extra_charge_on_statement
+Why is my money not in my account. I have already sent it out.,balance_not_updated_after_cheque_or_cash_deposit
+Let me know what the steps for the identity checks are,verify_my_identity
+I need my card as quick as possible,card_delivery_estimate
+In what increments can I top-up my card?,automatic_top_up
+When do i activate auto top-up?,automatic_top_up
+Do you charge for sending out more cards?,getting_spare_card
+Why am I being asked to verify my identity?,why_verify_identity
+What currencies can I use?,supported_cards_and_currencies
+I withdrew cash and I think the exchange rate is wrong.,wrong_exchange_rate_for_cash_withdrawal
+"My top up hasn't gone through yet, why?",pending_top_up
+"Hello, I have a question concerning an unfamiliar fee that I notice on my account. I see that you guys charge for ATM withdrawal. Never notice this fee until now. Can you please explain?",cash_withdrawal_charge
+The exchange rate was wrong in the foreign country I got cash in.,wrong_exchange_rate_for_cash_withdrawal
+I am still waiting for a transfer to show,pending_transfer
+How is the exchange rate calculated?,exchange_rate
+"There was an extra fee when I paid with my card, why was i charged this extra fee?",card_payment_fee_charged
+My credit card seems to have been declined for top up. Why is it not going through? Can you tell me what's going on?,top_up_failed
+"My transfers keep on getting declined. My card was working fine up until now, however it has suddenly stopped working. Why is this?",declined_transfer
+How long does it take for my physical card to be delivered.,card_delivery_estimate
+I need to know why a money transfer is available.,pending_transfer
+I must make several disposable cards every day.,disposable_card_limits
+What do I do to link my new card?,card_linking
+Why is their a charge pending on my card still?,pending_card_payment
+Why is there extra cash in my account?,cash_withdrawal_not_recognised
+Can I get a refund on an item?,request_refund
+Can I get an update on my replacement card?,card_arrival
+How do I know what my PIN is?,get_physical_card
+In which countries can I get a card?,country_support
+Can I choose from either Visa or Mastercard?,visa_or_mastercard
+My card still hasn't been delivered,card_arrival
+I didn't make the direct debit payment on my account.,direct_debit_payment_not_recognised
+after i got married i need to change my name,edit_personal_details
+How can I tell the source for my available funds?,verify_source_of_funds
+I didn't get all the cash I asked for,wrong_amount_of_cash_received

data/gpai_initiatives/task.json

@@ -0,0 +1 @@
+{"name": "gpai_initiatives", "description": "", "data_columns": ["Name", "Link", "Organization / Author", "Brief Description", "Sector", "Geographical scope", "Target Audience", "Stage of Development", "Date started", "Country/region of origin", "Notes (including specific SDG(s) and OECD AI Principles addressed)"], "label_columns": {"Label: AI and Ethics": ["0", "1"], "Label: AI and Governance": ["0", "1"], "Label: AI and Social Good": ["0", "1"]}}

data/gpai_initiatives/train.csv

@@ -0,0 +1,64 @@
+Name,Link,Organization / Author,Brief Description,Sector,Geographical scope,Target Audience,Stage of Development,Date started,Country/region of origin,Notes (including specific SDG(s) and OECD AI Principles addressed),Label: AI and Ethics,Label: AI and Governance,Label: AI and Social Good
+AI Civic Forum,https://thefuturesociety.org/2020/04/30/the-ai-civic-forum-empowering-people-to-shape-the-future-of-ai/,"Algora Lab, University of Montreal, TFS, Mila","The AI Civic Forum (AICF) is a multi-stakeholder platform to proactively engage people around the world in discussions on AI ethics and governance. Anchored in a robust collective intelligence process, the objectives are delivered through four key deliverables: face-to-face deliberations; an online platform; an AI Literacy Toolkit and a Trainer-the-trainer Playbook. The AI Civic Forum is co-led by The Future Society, AlgoraLab and Mila. - The initiative's mission is: 'Bring together diverse communities of citizens, policymakers, academics, experts, private sector representatives and other key stakeholders to deliberate on the ethical design, deployment, and governance of AI.'",Mixed academia/non-profit,Global,"Civil Society, Policymakers, Citizens",Set-up in progress,June 2019,United States,,0,1,0
+Tieto’s AI ethics guidelines,https://www.tietoevry.com/contentassets/b097de43d84d4c84832f1fff2cb6a30d/tieto-s-ai-ethics-guidelines.pdf; https://www.tietoevry.com/en/newsroom/all-news-and-releases/press-releases/2018/10/tieto-strengthens-commitment-to-ethical-use-of-ai/,Tieto,"Foundation for all Tieto employees who develop or use AI in any capacity. Also trialling internal AI ethics course & certification, plus establishing new roles focusing on ethical values embedded within AI - The initiative's mission is: 'Part of a sustained drive to promote ethical AI product and service development across the company and advance ethical and responsible AI'",Private,Global,Employees,Published,Oct 2018,Finland,,1,1,0
+Charlevoix Common Vision for the Future of Artificial Intelligence,https://www.international.gc.ca/world-monde/international_relations-relations_internationales/g7/documents/2018-06-09-artificial-intelligence-artificielle.aspx?lang=eng,Leaders of the G7,"G7 leaders' vision for a predictable and stable policy environment that promotes innovation, as recognized by the 2018 G7 Montreal Ministerial Statement on Artificial Intelligence, and the multi-stakeholder, human-centric vision outlined in the 2017 G7 ICT and Industry Ministers’ Torino Declaration. - The initiative's mission is: 'To commit to 12 principles for the future of AI.'",International organisation,G7,self (gov),Published,Jun 2018,International,,1,1,0
+The National Artificial Intelligence Research and Development Strategic Plan,https://www.nitrd.gov/pubs/national_ai_rd_strategic_plan.pdf,National Science and Technology Council; Networking and Information Technology Research and Development Subcommittee,"Report laying out US national AI strategy - The initiative's mission is: 'High-level framework to identify scientific and technological needs in AI and track the progress and maximizing the impact of R&D investements to fill those needs; to establish priorities for federally-funded R&D in AI, looking beyong near-term AI capabiltites towards long-term transformational impacts of AI on society and the world.'",Public,US,Employees,Published,Oct 2016,United States,,1,0,0
+Humane-AI-Net - H2020-ICT-48 European network of AI excellence centres,https://www.humane-ai.eu/,European Commission with a network of 53 academic and industrial partners across Europe,"The HumanE AI Network leverages the synergies between the involved centers of excellence to develop the scientific foundations and technological breakthroughs needed to shape the AI revolution in a direction that is beneficial to humans both individually and societally, and that adheres to European ethical values and social, cultural, legal, and political norms. The core challenge is the development of robust, human-centred, trustworthy AI systems capable of what could be described as “understanding” humans, adapting to complex real-world environments, and appropriately interacting in complex social settings. The aim is to facilitate AI systems that enhance human capabilities and empower individuals and society as a whole while respecting human autonomy and self-determination. The HumanE AI Net project will engender the mobilization of a research landscape far beyond direct project funding, involve and engage European industry, reach out to relevant social stakeholders, and create a unique innovation ecosystem that provides a manyfold return on investment for the European economy and society.","Network of research centres of excellence in AI, academia and industries",European Union,"Research, industry, policy makers, citizens","Network design and research roadmap completed, operational phase started",2018,Germany,Initiative addressed to operationalize the full spectrum of OECD and European AI principles,1,0,1
+Guiding Principles on Trusted AI Ethics,https://www.teliacompany.com/globalassets/telia-company/documents/about-telia-company/public-policy/2018/guiding-principles-on-trusted-ai-ethics.pdf,Telia Company,"Guiding Principles to for operations and employees - The initiative's mission is: 'The proactive design, implementation, testing, use and followup of AI.'",Private,Global,Employees,Published,2019,Sweden,,1,1,0
+AI Explained - Non-technical  Guide for Policymakers,https://www.aiforpeace.org/library,AI for Peace,"A manual to explain AI basics to policymakers and interested individuals without technical background. - The initiative's mission is: 'Demystify what AI is, and demonstrate how it is already altering our lives and societies we live in.'",Civil society,Global,"Policymakers, Citizens",Published,February 2020,International,,0,1,1
+Dutch Artificial Intelligence Manifesto,http://ii.tudelft.nl/bnvki/wp-content/uploads/2018/09/Dutch-AI-Manifesto.pdf,"Special Interest Group on Artificial Intelligence (SIGAI), ICT Platform Netherlands (IPN)",Manifesto by the Dutch Special Interest Group on Artificial Intelligence - The initiative's mission is: 'In this manifesto the Special Interest Group on Artificial Intelligence (SIGAI) proposes a research agenda and identifies priorities that require investments to ensure AI research in the Netherlands is able to establish and maintain its leading role in the world. ',Academia,Netherlands,"multiple (Dutch government, researchers)",Published,September 2018,Netherlands,,1,1,0
+Declaración de Principios Éticos Para La IA de Latinoamérica,http://ia-latam.com/etica-ia-latam/,IA Latam,"A Declaration of Ethical Principles for AI in Latin America, created in line with IA Latam's committment to Responsible Innovation and Evolution - The initiative's mission is: 'To promote the creation of self-adhesion ethical criteria and frameworks that help IA Latam and guide everyone to continue for the best possible path always having as a goal the improvement of people's well-being and a better planet for new generation.'",Non-profit,Latin America,Public,Published,2019,Latin America,,1,1,0
+Oversight Board,https://www.oversightboard.com/news/announcing-the-first-members-of-the-oversight-board/,Oversight Board (for Facebook & Instagram's content moderation),"A group of 20 independent experts in charge of reviewing content brought to it by Facebook users or Facebook itself, to decide on whether the content is in accordance with Facebook's content policies and values.  - The initiative's mission is: 'Protect free expression by making principled, independent decisions about important pieces of content and by issuing policy advisory opinions on Facebook's content policies.'",Mixed,Global,Industry,Set-up in progress,January 2019,United States,,0,1,0
+Te Mana Raraunga (Māori Data Sovereignty Network),https://www.temanararaunga.maori.nz/kaupapa,Te Mana Raraunga (Māori Data Sovereignty Network,"Promotes the Māori data sovereignty through establishment of a charter on data governance and operations, expert database, advocacy, workshops, research, and other projects. - The initiative's mission is: 'The purpose of Te Mana Raraunga is to enable Māori Data Sovereignty and 
+to advance Māori aspirations for collective and individual wellbeing.'",Civil society,New Zealand,All,Publically launched,July 2015,New Zealand,,0,1,1
+"Science, Law and Society (SLS) Initiative - Principles for the Governance of AI",https://thefuturesociety.org/2017/07/15/principles-law-and-society-initiative/,The Future Society,"Principles for the Governance of AI - The initiative's mission is: 'To reap AI’s promise, while mitigating its risks.'",Non-profit,Global,"Policymakers, academics, technologists",Published,July 2017,United States,,1,0,0
+Opening the Black Box - Driving Business,https://www.canasean.com/wp-content/uploads/2019/08/XAI-White-Paper.pdf,Element AI,Explaining the business case for Explainable AI products and services such as those sold by.,Private,Global,Industry,Publically launched,August 2019,Canada,,0,0,1
+Countries & initiatives overview,https://oecd.ai/countries-and-initiatives,OECD Artificial Intelligence Policy Observatory,"Interactive database of AI policies and initiatives from countries, territories and other stakeholders - The initiative's mission is: 'To facilitate international co-operation, benchmarking and help develop best practices.'",International organisation,Global,All,Publically launched,2020,International,,0,1,0
+Stanford Human-Centered AI Initiative (HAI) - three principles,"https://hai.stanford.edu/blog/opening-gate#legacy_content:~:text=three%20principles%3A%20that%20we%20must%20study,subtle%20and%20nuanced%20as%20human%20intelligence.",Stanford HAI,"An interdisciplinary, global hub for AI learners, researchers, developers, builders and users from academia, government and industry, as well as leaders and policymakers who want to understand and influence AI’s impact and potential. - The initiative's mission is: 'To advance AI research, education, policy and practice to improve the human condition.'",Academia,US,"self (academics, citizens)",Published,March 2019,United States,,1,0,1
+Responsible bots: 10 guidelines for developers of conversational AI,https://www.microsoft.com/en-us/research/uploads/prod/2018/11/Bot_Guidelines_Nov_2018.pdf,Microsoft,Set of guidelines - The initiative's mission is: 'To help developers design bots that build trust in the company and service that the bot represents',Private,Global,developers,"Published v1 (""we fully expect that they will be revised over time
+in response to your feedback and our own experiences."")",Nov 2018,United States,,1,0,0
+"The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation",https://arxiv.org/ftp/arxiv/papers/1802/1802.07228.pdf,Future of Humanity Institute; University of Oxford; Centre for the Study of Existential Risk; University of Cambridge; Center for a New American Security; Electronic Frontier Foundation; OpenAI,"This report surveys the landscape of potential security threats from malicious uses of artificial intelligence technologies, and proposes ways to better forecast, prevent, and mitigate these threats. It includes four high level recommendations and high priority research areas. - The initiative's mission is: 'To forecast, prevent and mitigate the malicious use of AI'",Mixed academia/non-profit,Global,unspecified,Published,Feb 2018,United Kingdom,,1,1,0
+AI for Good Global Summit,https://aiforgood.itu.int/,International Telecommunications Union,Yearly summit in partnership with 35 UN agencies to foster discussion and coordination on fulfilling the Sustainable Development Goals using AI. - The initiative's mission is: 'Identify practical applications of AI and scale those solutions for global impact',Mixed,Global,All,Publically launched,2017,Switzerland,,0,0,1
+Artificial Intelligence Industry Code of Conduct (Consultation Version),https://www.secrss.com/articles/11099,Artificial Intelligence Industry Alliance,"Code of conduct for AI developers - The initiative's mission is: 'To promote the ethical self-discipline of China's artificial intelligence industry, build consensus, and promote the healthy development of artificial intelligence'",Professional association,China,All,Consultation published,2019,China,,1,0,0
+Partnership on AI Issue Area on Safety-Critical AI (SCAI),https://www.partnershiponai.org/tenets/,Partnership on AI,"Partnership on AI conducts research, organizes discussions, shares insights, provides thought leadership, consults with relevant third parties, responds to questions from the public and media, and creates educational material that advances the understanding of AI technologies including machine perception, learning, and automated reasoning. Their members have agreed to eight tenets. - The initiative's mission is: 'To help ensure AI benefits people and society.'",Private sector alliance,Global,self,Adopted,Sep 2016,International,,1,1,0
+Everyday Ethics for Artificial Intelligence. A practical guide for designers & developers,https://www.ibm.com/watson/assets/duo/pdf/everydayethics.pdf,IBM,"Report intended to ""begin a conversation"" about defining the ethics embedded in the desgn and development of AI systems. - The initiative's mission is: 'This guide provides discussion points concerning: Specific virtues that AI systems should possess; Guidance for designers and developers building and training AI.'",Private,Global,Employees and Civil Society,Published,2019,United States,,1,0,0
+"Work in the age of artificial intelligence. Four perspectives on the economy, employment, skills and ethics",https://julkaisut.valtioneuvosto.fi/handle/10024/160980,"Ministry of Economic Affairs and Employment, Finland","Collection of four main articles that discuss (1) the effects of artificial intelligence on general economic and employment trends; (2) the transformation of work and the labour market; (3) reforms on education and skills maintenance; and (4) ethics. - The initiative's mission is: 'Work out what happens to work/the economy in Finland given developments in AI, to turn artificial intelligence into a success factor for Finnish companies. Finland’s goal is to be a global leader in applying artificial intelligence'",Public,Finland,Government,Published,Sep 2018,Finland,,1,1,0
+You and AI,https://royalsociety.org/topics-policy/projects/machine-learning/you-and-ai/,The Royal Society; DeepMind,"A series of events, publications and multimedia content explaining machine learning, collecting views on machine learning and analyzing these views, run throughout 2018 by the Royal Society supported by DeepMind. - The initiative's mission is: 'Build greater understand of what machine learning and AI are, how the technology works, and the ways it may affect our lies.'",Mixed,United Kingdom,Citizens,Closed,2018,United Kingdom,,0,1,0
+The Toronto Declaration: Protecting the right to equality and non- discrimination in machine learning systems,https://www.torontodeclaration.org/,Access Now; Amnesty International,"The Toronto Declaration is a landmark statement on protecting human rights in the age of artificial intelligence. Led by Amnesty International and Access Now, the Declaration has been widely endorsed by the global human rights community. The Toronto Declaration proposes that human rights law and standards are put front and center in existing and emerging conversations and methods analyzing the impact of machine learning and related technologies. - The initiative's mission is: 'To serve as a useful resource for researchers, policy-makers and tech professionals looking for guidance on applying human rights principles and standards to new technologies.'",Mixed international organisation/non-profit,Canada,"multiple (states, private sector ac- tors)",Published,May 2018,International,,1,0,1
+Policy Recommendations on Augmented Intelligence in Health Care H-480.940,https://policysearch.ama-assn.org/policyfinder/detail/augmented%20intelligence?uri=%2FAMADoc%2FHOD.xml-H-480.940.xml,American Medical Association (AMA),"A policy to provide a broad framework for the evolution of AI in health care - The initiative's mission is: 'To ensure that the evolution of augmented intelligence (AI) in medicine benefits patients, physicians, and the health care community.'",Professional association,US,Medical professionals,Implemented,2018,United States,"AMA will seek to: 1. Leverage its ongoing engagement in digital health and other priority areas for improving patient outcomes and physicians’ professional satisfaction to help set priorities for health care AI. 2. Identify opportunities to integrate the perspective of practicing physicians into the development, design, validation, and implementation of health care AI. 3. Promote development of thoughtfully designed, high-quality, clinically validated health care AI that: a. is designed and evaluated in keeping with best practices in user-centered design, particularly for physicians and other members of the health care team; b. is transparent; c. conforms to leading standards for reproducibility; d. identifies and takes steps to address bias and avoids introducing or exacerbating health care disparities including when testing or deploying new AI tools on vulnerable populations; and e. safeguards patients’ and other individuals’ privacy interests and preserves the security and integrity of personal information. 4. Encourage education for patients, physicians, medical students, other health care professionals, and health administrators to promote greater understanding of the promise and limitations of health care AI. 5. Explore the legal implications of health care AI, such as issues of liability or intellectual property, and advocate for appropriate professional and governmental oversight for safe, effective, and equitable use of and access to health care AI.",0,1,0
+G20 Ministerial Statement on Trade and Digital Economy,https://www.mofa.go.jp/files/000486596.pdf,G20 Trade Ministers and Digital Economy Ministers,"Statement, principles and recommendations for the responsible stewardship of Trustworthy AI, resulting from 2 days meeting of G20 digital economy ministers - The initiative's mission is: 'To design and implement digital policies to maximize benefits and minimize the challenges from the development of the digital economy, and to overcome challenges with special attention to developing countries and underrepresented populations'",International organisation,Global,Policymakers,Published,2019,International,,1,1,0
+AI Repository,https://www.itu.int/en/ITU-T/AI/Pages/ai-repository.aspx,International Telecommunication Union (ITU),"Catalogue/inventory of AI initiatives which accelerate progress towards the “17 UN Sustainable Development Goals (SDGs)” - The initiative's mission is: 'To identify AI related projects, research initiatives, think-tanks and organizations that can accelerate progress towards the “17 UN Sustainable Development Goals (SDGs)”'",International organisation,Global,All,Publically launched,2020,Europe,,0,0,1
+AI Now 2017 Report,https://ainowinstitute.org/AI_Now_2017_Report.pdf,AI Now Institute,"Report which identifies new developments, emerging challenges and makes recommendations in four areas: labor and automation, bias and inclusion, rights and liberties, and ethics and governance - The initiative's mission is: 'To ensure that the benefits of AI will be shared broadly, and that risks can be identified and mitigated.'",Academia,Global,"multiple (core public agencies, companies, industry, universities, conferences, other stakeholders)",Published,Oct 2017,United States,,1,0,0
+"SoBigData (2015-2024, H2020-Excellent Science Research Infrastructures) Integrated Infrastructure for Social Mining & Big Data Analytics",http://www.sobigdata.eu/,"European Commission, Italian National research Council and 32 academic and industrial partners from 12 EU countries","SoBigData (2015-2024, H2020-Excellent Science Research Infrastructures) Integrated Infrastructure for Social Mining & Big Data Analytics. A research infrastructure for open data science for social good, at the second stage of “Advanced community”, aggregating 32 partners of 12 EU Countries. 
+
+SoBigData++ is a resource for sharing datasets, methods, research skills and computational resources for supporting the comprehension of social phenomena through the lens of BigData and Artificial Intelligence. Its uniqueness lies in its ability to make the effort of data science and artificial intelligence scientific communities accessible to multiple stakeholders on a unique ecosystem.  SoBigData++ is a distributed and multi-disciplinary RI aimed at using social mining, big data and artificial intelligence to understand our globally-interconnected society. The RI’s service platform empowers researchers for the design and execution of large-scale social mining experiments. Pushing the FAIR (findable, accessible, Interoperable, responsible)  and FACT (Fair, Accountable, Confidential and Transparent) principles, the RI  renders social mining experiments more efficiently designed, and repeatable by leveraging concrete tools that operationalize ethics, incorporating values and norms for privacy, fairness, transparency and pluralism. SoBigData++ RI is based on three pillars: 1) a data ecosystem for procurement, access, and data curation. 2) a platform of interoperable social data mining tools, and methodologies. Over 200 resources are available: curated data sets, analytical tools and services from six thematic clusters: text and social media mining; social network analysis; human mobility analytics; web analytics; visual analytics and social data. 3) a community comprising scientific, industrial and third-party stakeholders, such as policymakers, it counts  6,000 users registered to the e-infrastructure.
+
+SoBigData++ operates in six domains: Societal debates and online misinformation studies how the information is created and how it influences debates; Sustainable cities for citizens focuses on urban challenges, aiming at value creation for city stakeholders; Demography, economics and finance 2.0 measures the patterns of well-being and poverty at a local and global scale; Migration Studies analyses the development of new methods for studying migration phenomenon; Sports data science combines data with analytical tools to unveil the complexity underlying sports; Social Impacts of AI provides a forum for studying impacts of artificial intelligence on our society.  Finally, a special interest group is dedicated to Network Medicine aimed at combining Big data, omics, and network science to advance the frontiers of medicine
+
+SoBigData++ integrates 31 key European data science and AI centres. SoBigData++ combines research areas related to big data analytics, computational social science, digital humanities, city planners, and human-centred artificial intelligence. The project includes competences on ethics, economic analysis, media ecology, contrast to misinformation, environmental sustainability and energy saving.
+
+SoBigData++ works in synergy with different H2020 projects and network of excellence: HumaneAI-net, TAILOR, AI4EU, WeVerify, Pericles, NoBias, Clarin, and five SoBigData scientists are PIs of ERC grants in artificial intelligence. ","Research infrastructure provided by (mostly public) research centres, universities, but also industries and start-ups","Europe, Global","Multiple stakeholders: interdisciplinary researchers, innovators, industry, policy makers, citizens","Advanced integrated research infrastructure, operational since 2015",2015,Italy,"SoBigData++ is aimed at operationalizing the ethical principles set by Europe and OECD within the research infrastructure, using ethics-by-design approaches in the big data/AI projects developed in the ""exploratories"", research & innovation labs oriented at multiple SGD's related societal impact (sustainable cities for citizens, well-being, sustainable growth, migration studies, ecology of media and social media).",0,0,1
+AI factsheets 360,https://aifs360.mybluemix.net/,IBM,Template for capturing relevant information about the creation and deployment of an AI model or service - The initiative's mission is: 'To foster trust in AI by increasing transparency and enabling governance.',Private,Global,Developers,Published,2019,United States,,0,1,0
+Automated and Connected Driving: Report,https://www.bmvi.de/SharedDocs/EN/publications/report-ethics-commission-automated-and-connected-driving.pdf?__blob=publicationFile,"Federal Ministry of Transport and Digital Infrastructure, Ethics Commission, Germany","20 Ethical rules for automated and connected vehicular traffic - The initiative's mission is: 'What technological development guidelines are required to ensure that we do not blur the contours of a human society that places individuals, their freedom of development, their physical and intellectual integrity and their entitlement to social respect at the heart of its legal regime?'",Public,Germany,Government,Published,June 2017,Germany,,1,0,0
+Mozilla Voice STT (formerly Deep Speech),https://voice.mozilla.org/stt.html,Mozilla Foundation,"An open-source software project to build Speech To Text engine, with a community of developers, companies and researchers contributing to its growth. - The initiative's mission is: 'Make speech recognition technology openly available to developers.'",Non-profit,Global,Developers,Pilot,2018,United States,,0,0,1
+Recommendation on the Ethics of Artificial Intelligence & AI Decision Makers' Toolkit,https://en.unesco.org/artificial-intelligence/ethics,UNESCO's AHEG (Ad Hoc Expert Group),"Outline of ethical benefits, risks and recommendations around AI - The initiative's mission is: 'To guide the development and application of AI in a human-centred approach and to be trustworthy.'",International organisation,Global,All,Published,May 2020,International,,1,1,0
+Ethics Guidelines for Trustworthy AI,https://ec.europa.eu/digital-single-market/en/news/ethics-guidelines-trustworthy-ai,High-Level Expert Group on Artificial Intelligence,Proposes seven requirements that AI systems should meet in order to be deemed trustworthy - The initiative's mission is: 'To promote Trustworthy AI',International organisation,Europe,All,"Published, sought fedback",Apr 2019,Europe,,1,1,0
+Closing the Human Rights Gap in AI Governance,https://s3.amazonaws.com/element-ai-website-bucket/whitepaper-closing-the-human-rights-gap-in-ai-governance.pdf,Element AI,"A 20-page report explaining what investors, companies and government can do to implement due diligence, data governance an human rights by design practices into their activity.",Private,Canada,"Industry, Policymakers",Published,November 2019,Canada,,0,1,0
+Universal Guidelines for Artificial Intelligence,https://thepublicvoice.org/ai-universal-guidelines/,The Public Voice,"Twelve Universal Guidelines to inform and improve the design and use of AI. The Guidelines are intended to maximize the benefits of AI, to minimize the risk, and to ensure the protection of human rights. These Guidelines should be incorporated into ethical standards, adopted in national law and international agreements, and built into the design of systems. - The initiative's mission is: 'To inform and improve the design and use of A'",Mixed,Global,"multiple (institu- tions, govern- ments)",Published,Oct 2018,International,,1,1,0
+Seeking Ground Rules for AI,https://www.nytimes.com/2019/03/01/business/ethical-ai-recommendations.html,New York Times’ New Work Summit,10 recommendations to encourage the ethical use of AI - The initiative's mission is: 'To make AI trustworthy',Private,Global,All,Published,2019,United States,,1,1,0
+Principles for the Cognitive Era,https://www.ibm.com/blogs/think/2017/01/ibm-cognitive-principles/,IBM,"Describes three principles for AI development/deployment - The initiative's mission is: 'Guiding what IBM develops and brings to the world, and how they do so, to achieve the economic and societal potential of a cognitive future.'",Private,Global,Employees,Published,2017,United States,,1,0,0
+Everyday Ethics for AI,https://www.ibm.com/design/ai/ethics/everyday-ethics/,IBM,"Everyday ethics for AI provides discussion points concerning:
+    specific virtues that AI systems should possess;
+    guidance for designers and developers training and building AI.",Private,United States,Developers & deployers,Published,2014,United States,,1,1,0
+Discussion Paper: National Strategy for Artificial Intelligence,https://niti.gov.in/national-strategy-artificial-intelligence,National Institution for Transforming India (Niti Aayog),"National strategy for government of India. It is termed #AIForAll as it is focused on leveraging AI for inclusive growth in line with the Government policy of Sabka Saath Sabka Vikas. Role of the Government has been clearly delineated to develop the research ecosystem, promote adoption and address skilling challenges. The strategy also flags important issues like ethics, bias and privacy issues relating to AI and envisions Government promoting research in technology to address these concerns. The focus is on sectors like agriculture, health and education where public investment and lead would be necessary. - The initiative's mission is: 'Explore how India can leverage the transformative technologies to ensure social and inclusive growth in line with the development philosophy of the government.'",Public,India,Government,Published,June 2018,India,,1,0,0
+FRR Quality Mark,https://responsiblerobotics.org/quality-mark/,Fondation for Responsible Robotics,"Label on robotics products indicating a positive assessment from an independent external expert group on responsible robotics. This group takes into account security, safety, privacy, fairness, sustainability, accountability and transparency. - The initiative's mission is: 'empower innovators and producers to create their products in an ethical and responsible way'",Civil society,Global,Industry,Pilot,December 2015,Netherlands,,0,1,0
+Towards the deployment of smart virtual systems for emotional support to women suffering breast cancer,"Not yet. For more info, contact matias@cs.cinvestav.mx",Centro de Investigación y de Estudios Avanzados del Instituto Politécnico Nacional. CINVESTAV - IPN. Dr. Matías Alvarado,"Breast cancer is a hard traumatic event in women’s life. Each woman's personality influences the coping strategies used to deal with it. This project analyzes the emotional regulation by regarding the influence of personality in strategic choices. Emotions include  fear, anguish, sadness, depression, braveness, acceptance, among others. The psychology ""Big Five Model"" backs the analysis on personalities, namely conscious, open, agreeable, extrovert and neurotic. The strategies are: situation selection, modify situation, attention deployment, cognitive change and response modulation. The project establishes a quantitative correlation between personality/strategies/emotion-regulation. The analytical result may evolve in smart virtual systems for emotional support to women suffering breast cancer. The increasing world health problem of breast cancer makes the deployment of these kind of system more useful and relevant.",Academia,Local - Mexico,Health and medical stakeholders; to support women living with breast cancer.,Implementation,2020,Mexico,SDG 2 Health and well being,0,0,1
+"Building safe artificial intelligence: specification, robustness, and assurance",https://medium.com/@deepmindsafetyresearch/building-safe-artificial-intelligence-52f5f75058f1,DeepMind,"Blog post discussing three areas of technical AI safety - specification, robustness and assurance - with the goal of contributing to the development of the field, and encouraging substantive engagement with the technical ideas it discusses. 
+",Private,Gllobal,Researchers,Published,"September, 2018",United Kingdom,,0,1,0
+Detailed Explanation on Key Points Concerning AI Utilization Principles,https://www.soumu.go.jp/main_content/000658286.pdf,"The Conference toward AI Network Society, MIC",Manual for how to translate the AI Utilization Principles in practice,Public,Japan,Private Sector,Published,"9 August, 2019",Japan,,0,1,0
+Report on Artificial Intelligence and Human Society (Unofficial translation),https://www8.cao.go.jp/cstp/tyousakai/ai/summary/aisociety_en.pdf,Advisory Board on Artificial Intelligence and Human Society (initiative of the Minister of State for Science and Technology Policy),"Report exploring the influence of AI technologies on society to ensure these technologies are used safely and beneficially - The initiative's mission is: 'Discuss the influence of AI technologies on society to ensure these technologies are used safely and beneficially, to clarify what benefits are expected, what issues are to be concerned, what issues are to be resolved, and what attitudes are beneficial.'",Public,Japan,"multiple (researchers, government, businesses, public, educators)",Published,March 2017,Japan,,1,1,0
+Governing Artificial Intelligence. Upholding Human Rights & Dignity,https://datasociety.net/library/governing-artificial-intelligence/,Data & Society,Report which draws the connections between AI and human rights; reframes recent AI-related controversies through a human rights lens; and reviews current stakeholder efforts at the intersection of AI and human rights. - The initiative's mission is: 'Help target audience incorporate human rights into social and organizational contexts related to the development and governance of AI.',Non-profit,Global,"Technology companies, governments, intergovernmental organizations, civil society groups, academia",Published,Oct 2018,United States,,1,0,1
+ISO/IEC JTC 1/SC 42,https://www.iso.org/committee/6794475.html,International Standards Organisation,A group of 8 project working groups to standardize technologies in the area of Artificial Intelligence. - The initiative's mission is: 'Trustworthy AI',Mixed,Global,Industry,Published,2017,United States,,0,1,0
+Ethics Policy for Peaceful R&D,https://www.iiim.is/ethics-policy/,Icelandic Institute for Intelligent Machines (IIIM),"Takes aim at two major threats to societal prosperity and peace. On the one hand, increases in military spending continue throughout the world, including automated weapons development. Justified by “growing terrorist threats”, these actions are themselves resulting in increased use of undue and unjustified force, military and otherwise — the very thing they are aiming to suppress. On the other, the increased possibility — and in many cases clearly documented efforts — of governments wielding advanced technologies to spy on their law-abiding citizens, in numerous ways, and sidestep long-accepted public policy intended to protect private lives from public exposure has gradually become too acceptable. In the coming years and decades artificial intelligence (AI) technologies — and powerful automation in general — has the potential to make these matters significantly worse. - The initiative's mission is: 'Establishing an Ethical Policy for all current and future activities of IIIM'",Non-profit,Iceland,Employees,Published,N.d.,Iceland,,1,0,0
+ITI AI Policy Principles,https://www.itic.org/news-events/news-releases/iti-unveils-first-industry-wide-artificial-intelligence-policy-principles,Information Technology Industry Council (ITI),"Set of principles developed by ITI member companies. ITI is a trusted leader in  innovation policy that drives sustainable, ethical, and equitable growth and opportunity for all. - The initiative's mission is: 'To ensure that member companies allow AI to flourish while guarding against unwanted impacts, our industry is committing to a set of principles to formalize our promise to responsible design of AI. '",Private,Global,self (members),Published,Oct 2017,United States,,1,0,0
+Mid- to Long-Term Master Plan in Preparation for the Intelligent Information Society,https://english.msit.go.kr/cms/english/pl/policies2/__icsFiles/afieldfile/2017/07/20/Master%20Plan%20for%20the%20intelligent%20information%20society.pdf,Government of the Republic of Korea,"The report considers the role of AI alongside other converging technologies such as the Internet of Things, cloud computing, big data analysis, and mobile technologies. The report discusses a range of implications of AI related to the workforce and economy as well as lifestyles and living environments. It outlines core factors of success including growing the economy, providing opportunity to all, and improving everyone’s safety and happiness. It also lays out a “National Vision” which is “Realizing a Human-Centered Intelligent Information Society.” - The initiative's mission is: 'To foster an intelligent information society on the basis of public-private partnership, with businesses and citizens playing leading roles and the government and research community providing support.'",Public,South Korea,Government,Published,July 2017,South Korea,,1,0,0

data/medical_subdomain_of_clinical_notes/task.json

@@ -0,0 +1 @@
+{"name": "medical_subdomain_of_clinical_notes", "description": "", "data_columns": ["Note"], "label_columns": {"Label": ["cardiology", "gastroenterology", "nephrology", "neurology", "psychiatry", "pulmonary disease"]}}

data/neurips_impact_statement_risks/task.json

@@ -0,0 +1 @@
+{"name": "neurips_impact_statement_risks", "description": "", "data_columns": ["Paper title", "Paper link", "Impact statement"], "label_columns": {"Label": ["doesn't mention a harmful application", "mentions a harmful application"]}}

data/neurips_impact_statement_risks/train.csv

@@ -0,0 +1,51 @@
+Paper title,Paper link,Impact statement,Label
+Auto-Panoptic: Cooperative Multi-Component Architecture Search for Panoptic Segmentation,https://proceedings.neurips.cc/paper/2020/file/ec1f764517b7ffb52057af6df18142b7-Paper.pdf,"This work makes the first attempt to search for all key components of panoptic pipeline and manages to accomplish this via the proposed Cooperative Multi-Component Architecture Search and efficient Path-Priority Search Policy. Most related work in the literature of NAS for fine-grained vision tasks concentrates on searching a specific part of the network and the balance of the overall network is largely ignored. Nevertheless, this type of technology is essential to improve the upper bound of popular detectors and segmentation networks. This may inspire new work towards the efficient search of the overall architecture for fine-grained vision tasks, e.g., object detection, semantic segmentation, panoptic segmentation and so on. We are not aware of any imminent risks of placing anyone at a disadvantage. In the future, more constraints and optimization algorithms can be applied to strike the optimal trade-off between accuracy and latency to deliver customized architecture for different platforms and devices.",doesn't mention a harmful application
+Design Space for Graph Neural Networks,https://proceedings.neurips.cc/paper/2020/file/c5c3d4fe6b2cc463c7d7ecba17cc9de7-Paper.pdf,"Impact on GNN research . Our work brings in many valuable mindsets to the field of GNN research. For example, we fully adopt the principle of controlling model complexity when comparing different models, which is not yet adopted in most GNN papers. We focus on finding guidelines / principles when designing GNNs, rather than particular GNN instantiations. We emphasize that the best GNN designs can drastically differ across tasks (the state-of-the-art GNN model on one task may have poor performance on other tasks). We thus propose to evaluate models on diverse tasks measured by quantitative similarity metric. Rather than criticizing the weakness of existing GNN architectures, our goal is to build a framework that can help researchers understand GNN design choices when developing new models suitable for different applications. Our approach serves as a tool to demonstrate the innovation of a novel GNN model ( e.g. , in what kind of design spaces / task spaces, a proposed algorithmic advancement is helpful), or a novel GNN task ( e.g. , showing that the task is not similar to any existing tasks thus calls for new challenges of algorithmic development). Impact on machine learning research . Our approach is in fact applicable to general machine learning model design. Specifically, we hope the proposed controlled random search technique can assist fair evaluation of novel algorithmic advancements. To show whether a certain algorithmic advancement is useful, it is important to sample random model-task combinations, then investigate in what scenarios the algorithmic advancement indeed improves the performance. Additionally, the proposed task similarity metric can be used to understand similarities between general machine learning tasks, e.g. , classification of MNIST and CIFAR-10. Our ranking-based similarity metric is fully general, as long as different designs can be ranked by their performance. Impact on other research domains . Our framework provides an easier than ever support for experts in other disciplines to solve their problems via GNNs. Domain experts only need to provide properly formatted domain-specific datasets, then recommended GNN designs will be automatically picked and applied to the dataset. In the fastest mode, anchor GNN models will be applied to the novel task in order to measure its similarity with known GNN tasks, where the corresponding best GNN designs have been saved. Top GNN designs in the tasks with high similarity to the novel task will be applied. If computational resources permitted, a full grid search / random search over the design space can also be easily carried out to the new task. We believe this pipeline can significantly lower the barrier for applying GNN models, thus greatly promote the application of GNNs in other research domains. Impact on the society . As is discussed above, given its clarity and accessibility, we are confident that our general approach can inspire novel applications that are of high impact to the society. Additionally, its simplicity can also provide great opportunities for AI education, where students can learn from SOTA deep learning models and inspiring applications at ease.",doesn't mention a harmful application
+Learning the Geometry of Wave-Based Imaging,https://proceedings.neurips.cc/paper/2020/file/5e98d23afe19a774d1b2dcbefd5103eb-Paper.pdf,"We do not see any major ethical consequences of this work. Our work has implications in the fields of exploratory imaging — earthquake detection, medical imaging etc. Our work improves the quality and reliability of imaging in these fields. Improving these fields has direct societal impact in finding new natural preserves, improved diagnosis in healthcare etc. A failure of our system leaves machine learning unreliable in exploratory imaging. Our method provides strong out-of-distribution generalization and hence is not biased according to the data.",doesn't mention a harmful application
+Noise2Same: Optimizing A Self-Supervised Bound for Image Denoising,https://proceedings.neurips.cc/paper/2020/file/ea6b2efbdd4255a9f1b3bbc6399b58f4-Paper.pdf,"In this paper, we introduce Noise2Same, a self-supervised framework for deep image denoising. As Noise2Same does not need paired clean data, paired noisy data, nor the noise model, its application scenarios could be much broader than both traditional supervised and existing self-supervised denoising frameworks. The most direct application of Noise2Same is to perform denoising on digital images captured under poor conditions. Individuals and corporations related to photography may benefit from our work. Besides, Noise2Same could be applied as a pre-processing step for computer vision tasks such as object detection and segmentation [ 18], making the downstream algorithms more robust to noisy images. Also, specific research communities could benefit from the development of Noise2Same as well. For example, the capture of high-quality microscopy data of live cells, tissue, or nanomaterials is expensive in terms of budget and time [27]. Proper denoising algorithms allow researchers to obtain high-quality data from low-quality data and hence remove the need to capture high-quality data directly. In addition to image denoising applications, the self-supervised denoising framework could be extended to other domains such as audio noise reduction and single-cell [1]. On the negative aspect, as many imaging-based research tasks and computer vision applications may be built upon the denoising algorithms, the failure of Noise2Same could potentially lead to biases or failures in these tasks and applications.",mentions a harmful application
+When Counterpoint Meets Chinese Folk Melodies,https://proceedings.neurips.cc/paper/2020/file/bae876e53dab654a3d9d9768b1b7b91a-Paper.pdf,"The idea of integrating Western counterpoint into Chinese folk music generation is innovative. It would make positive broader impacts on three aspects: 1) It would facilitate more opportunities and challenges of music cultural exchanges at a much larger scale through automatic generation. For example, the inter-cultural style fused music could be used in Children’s enlightenment education to stimulate their interest in both cultures. 2) It would further the idea of collaborative counterpoint improvisation between two parts ( e . g ., a human and a machine) to music traditions where such interaction was less common. 3) The computer-generated music may “reshape the musical idiom”[23], which may bring more opportunities and possibilities to produce creative music. The proposed work may also have some potential negative societal impacts: 1) Similar to other computational creativity research, the generated music has the possibility of plagiarism by copying short snippets from the training corpus, even though copyright infringement is not a concern as neither folk melodies nor Bach’s music has copyright. That being said, our online music generation approach conditions music generation on past human and machine generation, and is less likely to directly copy snippets than offline approaches do. 2) The proposed innovative music generation approach may cause disruptions to current music professions, even deprive them of their means of existence[23]. However, it also opens new areas and creates new needs in this we-media era . Overall, we believe that the positive impacts significantly outweigh the negative impacts.",mentions a harmful application
+Learning from Label Proportions: A Mutual Contamination Framework,https://proceedings.neurips.cc/paper/2020/file/fcde14913c766cf307c75059e0e89af5-Paper.pdf,"LLP has been discussed as a model for summarizing a fully labeled dataset for public dissemination. The idea is that individual labels are not disclosed, so some degree of privacy is retained. As we show, consistent classification is still possible in this setting. If the two class-conditional distributions are nonoverlapping, labels of training instances can be recovered with no uncertainty by an optimal classifier. If the class-conditional distributions have some overlap, training instances in the nonoverlapping region can still be labeled with no uncertainty, while training instances in the overlapping regions can have their labels guessed with some uncertainty, depending on the degree of overlap.",doesn't mention a harmful application
+Limits to Depth Efficiencies of Self-Attention,https://proceedings.neurips.cc/paper/2020/file/ff4dfdf5904e920ce52b48c1cef97829-Paper.pdf,"Our work aims at providing fundamental guidelines which can assist all fields that employ Transformer-based architectures to use more efficient models. This way, these fields can achieve their goals while consuming less resources. Additionally, this work made an effort to provide a theoretical interpretation by examining the (many) empirical signals already published by others, while providing only a required minimum of further experimentation. This was done under the belief that while experiments are crucial for the advancement of the field, it is important not to conduct them superfluously as they incur an environmental price [Schwartz et al., 2019].",doesn't mention a harmful application
+Meta-Consolidation for Continual Learning,https://proceedings.neurips.cc/paper/2020/file/a5585a4d4b12277fee5cad0880611bc6-Paper.pdf,"(as required by NeurIPS 2020 CFP) Continual learning is a key desiderata for Artificial General Intelligence (AGI). Hence, this line of research has the benefits as well as the pitfalls of any other research effort geared in this direction. In particular, our work can help deliver impact on making smarter AI products and services, which can learn and update themselves on-the-fly when newer tasks and domains are encountered, without forgetting previously acquired knowledge. This is a necessity in any large-scale deployments of machine learning and computer vision, including in social media, e-commerce, surveillance, e- governance, etc - each of which have newer settings, tasks or domains added continually over time. Any negative effect of our work, such as legal and ethical concerns, are not unique to this work - to the best of our knowledge, but are shared with any other new development in machine learning, in general.",mentions a harmful application
+Learning to Incentivize Other Learning Agents,https://proceedings.neurips.cc/paper/2020/file/ad7ed5d47b9baceb12045a929e7e2f66-Paper.pdf,"Our work is a step toward the goal of ensuring the common good in a potential future where independent reinforcement learning agents interact with one another and/or with humans in the real world. We have shown that cooperation can emerge by introducing an additional learned incentive function that enables one agent to affect another agent’s reward directly. However, as agents still independently maximize their own individual rewards, it is open as to how to prevent an agent from misusing the incentive function to exploit others. One approach for future research to address this concern is to establish new connections between our work and the emerging literature on reward tampering [11]. By sparking a discussion on this important aspect of multi-agent interaction, we believe our work has a positive impact on the long-term research endeavor that is necessary for RL agents to be deployed safely in real-world applications.",doesn't mention a harmful application
+An Improved Analysis of (Variance-Reduced) Policy Gradient and Natural Policy Gradient Methods,https://proceedings.neurips.cc/paper/2020/file/56577889b3c1cd083b6d7b32d32f99d5-Paper.pdf,"The results of this paper improves the performance of policy-gradient methods for reinforcement learning, as well as our understanding to the existing methods. Through reinforcement learning, our study will also benefit several research communities such as machine learning and robotics. We do not believe that the results in this work will cause any ethical issue, or put anyone at a disadvantage in our society.",doesn't mention a harmful application
+Sample-Efficient Reinforcement Learning of Undercomplete POMDPs,https://proceedings.neurips.cc/paper/2020/file/d783823cc6284b929c2cd8df2167d212-Paper.pdf,"As this is a theoretical contribution, we do not envision that our direct results will have a tangible societal impact. Our broader line of inquiry could impact a line of thinking in a way that provides additional means to provide confidence intervals relevant for planning and learning. There is an increasing needs for applications to understand planning under uncertainty in the broader context of safety and reliability, and POMDPs provide one potential framework.",doesn't mention a harmful application
+Reward-rational (implicit) choice: A unifying formalism for reward learning,https://proceedings.neurips.cc/paper/2020/file/2f10c1578a0706e06b6d7db6f0b4a6af-Paper.pdf,"As AI capability advances, it is becoming increasingly important to align the objectives of AI agents to what people want. From how assistive robots can best help their users, to how autonomous cars should trade off between safety risk and efficiency, to how recommender systems should balance revenue considerations with longer-term user happiness and with avoiding influencing user views, agents cannot rely on a reward function specified once and set in stone. By putting different sources of information about the reward explicitly under the same framework, we hope our paper contributes towards a future in which agents maintain uncertainty over what their reward should be, and use different types of feedback from humans to refine their estimate and become better aligned with what people want over time – be them designers or end-users. On the flip side, changing reward functions also raises its own set of risks and challenges. First, the relationship between designer objectives and end-user objectives is not clear. Our framework can be used to adapt agents to end-users preferences, but this takes away control from the system designers. This might be desirable for, say, home robots, but not for safety-critical systems like autonomous cars, where designers might need to enforce certain constraints a-priori on the reward adaptation process. More broadly, most systems have multiple stake-holders, and what it means to do ethical preference aggregation remains an open problem. Further, if the robot’s model of the human is misspecified, adaptation might lead to more harm than good, with the robot inferring a worse reward function than what a designer could specify by hand.",mentions a harmful application
+Flows for simultaneous manifold learning and density estimation,https://proceedings.neurips.cc/paper/2020/file/051928341be67dcba03f0e04104d9047-Paper.pdf,"Manifold-learning flows have the potential to improve the efficiency with which scientists extract knowledge from large-scale experiments. Many phenomena have their most accurate description in terms of complex computer simulations which do not admit a tractable likelihood. In this common case, normalizing flows can be trained on synthetic data and used as a surrogate for the likelihood function, enabling high-quality inference on model parameters [21]. When the data have a manifold structure, manifold-learning flows may improve the quality and efficiency of this process further and ultimately contribute to scientific progress. We have demonstrated this with a real-world particle physics dataset, though the same technique is applicable to fields as diverse as neuroscience, systems biology, and epidemiology. All generative models carry a risk of being abused for the generation of fake data that are then masqueraded as real documents. This danger also applies to manifold-learning flows. While manifold-learning flows are currently far away from being able to generate realistic high-resolution images, videos, or audio, this concern should be kept in mind in the long term. Finally, the models we trained on image datasets of human faces clearly lack diversity. They reproduce and reinforce the biases inherent in the training data. Before using such (or other) models in any real-life application, it is crucial to understand, measure, and mitigate such biases.",mentions a harmful application
+Implicit Neural Representations with Periodic Activation Functions,https://proceedings.neurips.cc/paper/2020/file/53c04118df112c13a8c34b38343b9c10-Paper.pdf,"The proposed SIREN representation enables accurate representations of natural signals, such as images, audio, and video in a deep learning framework. This may be an enabler for downstream tasks involving such signals, such as classification for images or speech-to-text systems for audio. Such applications may be leveraged for both positive and negative ends. SIREN may in the future further enable novel approaches to the generation of such signals. This has potential for misuse in impersonating actors without their consent. For an in-depth discussion of such so-called DeepFakes, we refer the reader to a recent review article on neural rendering [16].",mentions a harmful application
+Neural Message Passing for Multi-Relational Ordered and Recursive Hypergraphs,https://proceedings.neurips.cc/paper/2020/file/217eedd1ba8c592db97d0dbe54c7adfc-Paper.pdf,"Message Passing Neural Networks (MPNNs) are a framework for deep learning on graph structured data. Graph structures are universal and very generic structures commonly seen in various forms in computer vision, natural language processing, recommender systems, traffic prediction, generative models, and many more. Graphs can have many variations such as multi-relational, heterogeneous, hypergraphs, etc. Our research in this paper unifies several existing MPNN methods on these variations. While we show how our research could be used for academic networks, and factual knowledge, it opens up many more possibilities in natural language processing (NLP). We see opportunities for research applying our work for beneficial puroposes, such as investigating whether we could improve performance of NLP tasks such as machine reading comprehension, relation extraction, machine translation, and many more. Potentially hazardous applications include trying to predict criminality or credit from social networks. Such applications may reproduce and exacerbate bias and readers of the paper should be aware that the presented model should not applied naively to such tasks.",mentions a harmful application
+COT-GAN: Generating Sequential Data via Causal Optimal Transport,https://proceedings.neurips.cc/paper/2020/file/641d77dd5271fca28764612a028d9c8e-Paper.pdf,"The COT-GAN algorithm introduced in this paper is suitable to generate sequential data, when the real dataset consists of i.i.d. sequences or of stationary time series. It opens up doors to many applications that can benefit from time series synthesis. For example, researchers often do not have access to abundant training data due to privacy concerns, high cost, and data scarcity. This hinders the capability of building accurate predictive models. Ongoing research is aimed at developing a modified COT-GAN algorithm to generate financial time series. The high non-stationarity of financial data requires different features and architectures, whilst causality when measuring distances between sequences remains the crucial tool. The application to market generation is of main interest for the financial and insurance industry, for example in model- independent pricing and hedging, portfolio selection, risk management, and stress testing. In broader scientific research, our approach can be used to estimate from data the parameters of simulation-based models that describe physical processes. These models can be, for instance, differential equations describing neural activities, compartmental models in epidemiology, and chemical reactions involving multiple reagents.",doesn't mention a harmful application
+Cream of the Crop: Distilling Prioritized Paths For One-Shot Neural Architecture Search,https://proceedings.neurips.cc/paper/2020/file/d072677d210ac4c03ba046120f0802ec-Paper.pdf,"Similar to previous NAS works, this work does not have immediate societal impact, since the algorithm is only designed for image classification, but it can indirectly impact society. As an example, our work may inspire the creation of new algorithms and applications with direct societal implications. Moreover, compared with other NAS methods that require additional teacher model to guide the training process, our method does not need any external teacher models. So our method can be used in a closed data system, ensuring the privacy of user data.",doesn't mention a harmful application
+Deep Evidential Regression,https://proceedings.neurips.cc/paper/2020/file/aab085461de182608ee9f607f3f7d18f-Paper.pdf,"Uncertainty estimation for neural networks has very significant societal impact. Neural networks are increasingly being trained as black-box predictors and being placed in larger decision systems where errors in their predictions can pose immediate threat to downstream tasks. Systematic methods for calibrated uncertainty estimation under these conditions are needed, especially as these systems are deployed in safety critical domains, such for autonomous vehicle control [29], medical diagnosis [43], or in settings with large dataset imbalances and bias such as crime forecasting [24] and facial recognition [3]. This work is complementary to a large portion of machine learning research which is continually pushing the boundaries on neural network precision and accuracy. Instead of solely optimizing larger models for increased performance, our method focuses on how these models can be equipped with the ability to estimate their own confidence. Our results demonstrating superior calibration of our method over baselines are also critical in ensuring that we can place a certain level of trust in these algorithms and in understanding when they say “I don’t know”. While there are clear and broad benefits of uncertainty estimation in machine learning, we believe it is also important to recognize potential societal challenges that may arise. With increased performance and uncertainty estimation capabilities, humans will inevitably become increasingly trusting in a model’s predictions, as well as its ability to catch dangerous or uncertain decisions before they are executed. Thus, it is important to continue to pursue redundancy in such learning systems to increase the likelihood that mistakes can be caught and corrected independently.",mentions a harmful application
+The Value Equivalence Principle for Model-Based Reinforcement Learning,https://proceedings.neurips.cc/paper/2020/file/3bb585ea00014b0e3ebe4c6dd165a358-Paper.pdf,"The bulk of the research presented in this paper consists of foundational theoretical results about the learning of models for model-based reinforcement learning agents. While applications of these agents can have social impacts depending upon their use, our results merely serve to illuminate desirable properties of models and facilitate the subsequent training of agents using them. In short, this work is largely theoretical and does not present any foreseeable societal impact, except in the general concerns over progress in artificial intelligence.",doesn't mention a harmful application
+Graph Policy Network for Transferable Active Learning on Graphs,https://proceedings.neurips.cc/paper/2020/file/73740ea85c4ec25f00f9acbd859f861d-Paper.pdf,"Graph-structured data are ubiquitous in real world, covering a variety of domains and applications such as social science, biology, medicine, and political science. In many domains such as biology and medicine, annotating a large number of labeled data could be extremely expensive and time consuming. Therefore, the algorithm proposed in this paper could help significantly reduce the labeling efforts in these domains — we can train systems on domains where labeled data are available, then transfer to those lower-resource domains. We believe such systems can help accelerating some research and develop processes that usually take a long time, in domains such as drug development. It can potentially also lower the cost for such research by reducing the need of expert-annotations. However, we also acknowledge potential social and ethical issues related to our work. 1. Our proposed system can effectively reduce the need of human annotations. However, in a broader point of view, this can potentially lead to a reduction of employment opportunities which may cause layoff to data annotators. 2. GNNs are widely used in domains related to critical needs such as healthcare and drug development. The community needs to be extra cautious and rigorous since any mistake may cause harm to patients. 3. Training the policy network for active learning on multiple graphs is relatively time - and computational resource - consuming. This line of research may produce more carbon footprint compared to some other work. Therefore, how to accelerate the training process by developing more efficient algorithms requires further investigation. Nonetheless, we believe that the directions of active learning and transfer learning provide a hopeful path towards our ultimate goal of data efficiency and interpretable machine learning.",mentions a harmful application
+User-Dependent Neural Sequence Models for Continuous-Time Event Data,https://proceedings.neurips.cc/paper/2020/file/f56de5ef149cf0aedcc8f4797031e229-Paper.pdf,"While many of the successful and highly-visible applications of machine learning are in classification and regression, there are a broad range of applications that don’t naturally fit into these categories and that can potentially benefit significantly from machine learning approaches. In particular, in this paper we focus on continuous-time event data, which is very common in real-world applications but has not yet seen significant attention from the ML research community. There are multiple important problems in society where such data is common and that could benefit from the development of better predictive and simulation, including: • Education: Understanding of individual learning habits of students, especially in online educa- tional programs, could improve and allow for more personalized curricula. • Medicine: Customized tracking and predictions of medical events could save lives and improve patients’ quality of living. • Behavioral Models: Person-specific simulations of their behavior can lead to better systematic understandings of people’s social activities and actions in day-to-day lives. • Cybersecurity: Through the user identification capabilities, our work could aid in cyber-security applications for the purposes of identifying fraud detection and identify theft. Another potential positive broad impact of the work, is that by utilizing amortized VI, our methods do not require further costly training or fine-tuning to accommodate new users, which can potentially produce energy savings and lessen environmental impact in a production setting. On the other hand, as with many machine learning technologies, there is also always the potential for negative impact from a societal perspective. For example, more accurate individualized models for user-generated data could be used in a negative fashion for applications such as surveillance (e.g., to monitor and negatively impact individuals in protected groups). In addition, better predictions and recommendations for products and services, through explicitly conditioning on prior behavior from a user, could potentially further worsen existing privacy concerns.",mentions a harmful application
+Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID,https://proceedings.neurips.cc/paper/2020/file/821fa74b50ba3f7cba1e6c53e8fa6845-Paper.pdf,"Our method can help to identify and track different types of objects ( e . g ., vehicles, cyclists, pedestrians, etc . ) across different cameras (domains), thus boosting the development of smart retail, smart transportation, and smart security systems in the future metropolises. In addition, our proposed self-paced contrastive learning is quite general and not limited to the specific research field of object re-ID. It can be well extended to broader research areas, including unsupervised and semi-supervised representation learning. However, object re-ID systems, when applied to identify pedestrians and vehicles in surveillance systems, might give rise to the infringement of people’s privacy, since such re-ID systems often rely on non-consensual surveillance data for training, i . e ., it is unlikely that all human subjects even knew they were being recorded. Therefore, governments and officials need to carefully establish strict regulations and laws to control the usage of re-ID technologies. Otherwise, re-ID technologies can potentially equip malicious actors with the ability to surveil pedestrians or vehicles through multiple CCTV cameras without their consent. The research committee should also avoid using the datasets with ethics issues, e . g ., DukeMTMC [37], which has been taken down due to the violation of data collection terms, should no longer be used. We would not evaluate our method on DukeMTMC related benchmarks as well. Furthermore, we should be cautious of the misidentification of the re-ID systems to avoid possible disturbance. Also, note that the demographic makeup of the datasets used is not representative of the broader population.",mentions a harmful application
+Real World Games Look Like Spinning Tops,https://proceedings.neurips.cc/paper/2020/file/ca172e964907a97d5ebd876bfdd4adbd-Paper.pdf,"This work focuses on better understanding of mathematical properties of real world games and how they could be used to understand successful AI techniques that were developed in the past. Since we focus on retrospective analysis of a mathematical phenomenon, on exposing an existing structure, and deepening our understanding of the world, we do not see any direct risks it entails. Introduced notions and insights could be used to build better, more engaging AI agents for people to play with in real world games (e.g. AIs that grow with the player, matching their strengths and weaknesses). In a broader spectrum, some of the insights could be used for designing and implementing new games, that humans would fine enjoyable though challenges they pose. In particular it could be a viewed as a model for measuring how much notion of progress the game consists of. However, we acknowledge that methods enabling improved analysis of games may be used for designing products with potentially negative consequences (e.g., games that are highly addictive) rather than positive (e.g., games that are enjoyable and mentally developing).",mentions a harmful application
+Adapting Neural Architectures Between Domains,https://proceedings.neurips.cc/paper/2020/file/08f38e0434442128fab5ead6217ca759-Paper.pdf,This paper provides a novel perspective of cross-domain generalization in neural architecture search towards the efficient design of neural architectures with strong generalizability. This will lead to a better understanding of the generalizability of neural architectures. The proposed method will be used to design neural architectures for computer vision tasks with affordable computation cost.,doesn't mention a harmful application
+Modeling Noisy Annotations for Crowd Counting,https://proceedings.neurips.cc/paper/2020/file/22bb543b251c39ccdad8063d486987bb-Paper.pdf,"In this paper, we introduce a novel loss function for counting crowd numbers by explicitly considering annotation noise. It can be applied to any density map based network architecture and improve the counting accuracy generally. The research is also helpful for monitoring the crowd number in public and prevent the accidents caused by overcrowding. It could also be used in retail businesses to estimate the occupancy of a store or area, which helps with personal and resource management. Our method could also be applied to other objects, such as cell counting, plant/animal counting, etc, and other research areas that use point-wise annotations, e.g., eye gaze estimation. Since the research is based on images captured by cameras, users may be concerned about the privacy problem. However, our method does not directly detect or track individuals, and thus this concern may be eased.",doesn't mention a harmful application
+Byzantine Resilient Distributed Multi-Task Learning,https://proceedings.neurips.cc/paper/2020/file/d37eb50d868361ea729bb4147eb3c1d8-Paper.pdf,"The problem of Byzantine resilient aggregation of distributed machine learning models has been actively studied in recent years; however, the issue of Byzantine resilient distributed learning in multi-task networks has received much less attention. It is a general intuition that MTL is robust and resilient to cyber-attacks since it can identify attackers by measuring similarities between neighbors. In this paper, we have shown that some commonly used similarity measures are not resilient against certain attacks. With an increase in data heterogeneity, we hope this work could highlight the security and privacy concerns in designing distributed MTL frameworks.",doesn't mention a harmful application
+From Predictions to Decisions: Using L kahead Regularization,https://proceedings.neurips.cc/paper/2020/file/2adcfc3929e7c03fac3100d3ad51da26-Paper.pdf,"In our work, the learning objective was designed to align with and support the possible use of a predictive model to drive decisions by users. It is our belief that a responsible and transparent deployment of models with “lookahead-like"" regularization components should avoid the kinds of mistakes that can be made when predictive methods are conflated with causally valid methods. At the same time, we have made a strong simplifying assumption, that of covariate shift, which requires that the relationship between covariates and outcome variables is invariant as decisions are made and the feature distribution changes. This strong assumption is made to ensure validity for the lookahead regularization, since we need to be able to perform inference about counterfactual observations. As discussed by Mueller et al. [ 31] and Peters et al. [34], there exist real-world tasks that reasonably satisfy this assumption, and yet at the same time, other tasks— notably those with unobserved confounders —where this assumption would be violated. Moreover, this assumption is not testable on the observational data. This, along with the need to make an assumption about the user decision model, means that an application of the method proposed here should be done with care and will require some domain knowledge to understand whether or not the assumptions are plausible. Furthermore, the validity of the interval estimates requires that any assumptions for the interval model used are satisfied and that weights w provide a reasonable estimation of p /p . In particular, fitting to p which has little to no overlap with p (see Figure 2) may result in underestimating the possibility of bad outcomes. If used carefully and successfully, then the system provides safety and protects against the misuse of a model. If used in a domain for which the assumptions fail to hold then the framework could make things worse, by trading accuracy for an incorrect view of user decisions and the effect of these decisions on outcomes. We would also caution against any specific interpretation of the application of the model to the wine and diabetes data sets. We note that model misspecification of f ∗ could result in arbitrarily bad outcomes, and estimating f ∗ in any high-stakes setting requires substantial domain knowledge and should err on the side of caution. We use the data sets for purely illustrative purposes because we believe the results are representative of the kinds of results that are available when the method is correctly applied to a domain of interest.",mentions a harmful application
+Finite-Time Analysis of Round-Robin Kullback-Leibler Upper Confidence Bounds for Optimal Adaptive Allocation with Multiple Plays and Markovian Rewards,https://proceedings.neurips.cc/paper/2020/file/597c7b407a02cc0a92167e7a371eca25-Paper.pdf,"This work touches upon a very old problem dating back to 1933 and the work of [39]. Therefore, we don’t anticipate any new societal impacts or ethical aspects, that are not well understood by now.",doesn't mention a harmful application
+Towards Interaction Detection Using Topological Analysis on Neural Networks,https://proceedings.neurips.cc/paper/2020/file/473803f0f2ebd77d83ee60daaa61f381-Paper.pdf,"The proposed PID algorithm can be applied in various fields because it provides knowledge about a domain. Any researcher who needs to design experiments might benefit from our proposed algorithm in the sense that it can help researchers formulate hypotheses that could lead to new data collection and experiments. For example, PID can help us discover the combined effects of drugs on human body: By utilizing PID on patients’ records, we might find using Phenelzine togther with Fluoxetine has a strong interaction effect towards serotonin syndrome. Thus, PID has great potential in helping the development of new therapies for saving lives. Also, this project will lead to effective and efficient algorithms for finding useful any-order crossing features in an automated way. Finding useful crossing features is one of the most crucial task in the Recommender Systems. Engineers and Scientists in E-commerce companies may benefit from our results that our algorithm can alleviate the human effect on finding these useful patterns in the data.",doesn't mention a harmful application
+Why Normalizing Flows Fail to Detect Out-of-Distribution Data,https://proceedings.neurips.cc/paper/2020/file/ecb9fe2fbb99c31f567e9823e884dbec-Paper.pdf,"Out-of-distribution detection is crucial for robust, reliable and fair machine learning systems. Mitchell et al. [27] and Gebru et al. [13] argue that applying machine learning models outside of the context where they were trained and tested can lead to dangerous and discriminatory outcomes in high-stake domains. We hope that our work will generally contribute to the understanding of out-of-distribution detection and facilitate methodological progress in this area.",doesn't mention a harmful application
+AdaShare: Learning What To Share For Efficient Deep Multi-Task Learning,https://proceedings.neurips.cc/paper/2020/file/634841a6831464b64c072c8510c7f35c-Paper.pdf,"Our research improves the capacity of deep neural networks to solve many tasks at once in a more efficient manner. It enables the use of smaller networks to support more tasks, while performing knowledge transfer between related tasks to improve their accuracy. For example, we showed that our proposed approach can solve five computer vision tasks (semantic segmentation, surface normal prediction, depth prediction, keypoint detection and edge estimation) with 80% fewer parameters while achieving the same performance as the standard approach. Our approach can thus have a positive impact on applications that require multiple tasks such as computer vision for robotics. Potential applications could be in assistive robots, autonomous navigation, robotic picking and packaging, rescue and emergency robotics and AR/VR systems. Our research can reduce the memory and power consumption of such systems and enable them to be deployed for longer periods of time and become smaller and more agile. The lessened power consumption could have a high impact on the environment as AI systems become more prevalent. Negative impacts of our research are difficult to predict, however, it shares many of the pitfalls associated with deep learning models. These include susceptibility to adversarial attacks and data poisoning, dataset bias, and lack of interpretablity. Other risks associated with deployment of computer vision systems include privacy violations when images are captured without consent, or used to track individuals for profit, or increased automation resulting in job losses. While we believe that these issues should be mitigated, they are beyond the scope of this paper. Furthermore, we should be cautious of the result of failure of the system which could impact the performance/user experience of the high-level AI systems relied on our research.",mentions a harmful application
+AOT: Appearance Optimal Transport Based Identity Swapping for Forgery Detection,https://proceedings.neurips.cc/paper/2020/file/f718499c1c8cef6730f9fd03c8125cab-Paper.pdf,"Deepfake refers to synthesized media in which a portrait of a person in real media is replaced by that of someone else. Deepfakes have been widely applied in the digital entertainment industry, but they also present potential threats to the public. Identity swapping is an approach to produce Deepfakes and is also the research direction of this paper. Given the sensitivity of Deepfakes and their potential negative impacts, we further discuss the potential threats and the corresponding mitigation solutions with respect to our work.",mentions a harmful application
+Permute-and-Flip: A new mechanism for differentially private selection,https://proceedings.neurips.cc/paper/2020/file/01e00f2f4bfcbb7505cb641066f2859b-Paper.pdf,"Our work fi ts in the established research area of differential privacy, which enables the positive societal bene fi ts of gleaning insight and utility from data sets about people while offering formal guarantees of privacy to individuals who contribute data. While these bene fi ts are largely positive, unintended harms could arise due to misapplication of differential privacy or misconceptions about its guarantees. Additionally, dif fi cult social choices are faced when deciding how to balance privacy and utility. Our work addresses a foundational differential privacy task and enables better utility-privacy tradeoffs within this broader context.",mentions a harmful application
+Classification with Valid and Adaptive Coverage,https://proceedings.neurips.cc/paper/2020/file/244edd7e85dc81602b7615cd705545f5-Paper.pdf,"Machine learning algorithms are increasingly relied upon by decision makers. It is therefore crucial to combine the predictive performance of such complex machinery with practical guarantees on the reliability and uncertainty of their output. We view the calibration methods presented in this paper as an important step towards this goal. In fact, uncertainty estimation is an effective way to quantify and communicate the benefits and limitations of machine learning. Moreover, the proposed methodologies provide an attractive way to move beyond the standard prediction accuracy measure used to compare algorithms. For instance, one can compare the performance of two candidate predictors, e.g., random forest and neural network (see Figure 3), by looking at the size of the corresponding prediction sets and/or their their conditional coverage. Finally, the approximate conditional coverage that we seek in this work is highly relevant within the broader framework of fairness, as discussed by [17] within a regression setting. While our approximate conditional coverage already implicitly reduces the risk of unwanted bias, an equalized coverage requirement [17] can also be easily incorporated into our methods to explicitly avoid discrimination based on protected categories. We conclude by emphasizing that the validity of our methods relies on the exchangeability of the data points. If this assumption is violated (e.g., with time-series data), our prediction sets may not have the right coverage. A general suggestion here is to always try to leverage specific knowledge of the data and of the application domain to judge whether the exchangeability assumption is reasonable. Finally, our data-splitting techniques in Section 4 offer a practical way to verify empirically the validity of the predictions on any given data set.",doesn't mention a harmful application
+Learning Kernel Tests Without Data Splitting,https://proceedings.neurips.cc/paper/2020/file/44f683a84163b3523afe57c2e008bc8c-Paper.pdf,"Hypothesis testing and valid inference after model selection are fundamental problems in statistics, which have recently attracted increasing attention also in machine learning. Kernel tests such as MMD are not only used for statistical testing, but also to design algorithms for deep learning and GANs [41, 42]. The question of how to select the test statistic naturally arises in kernel-based tests because of the kernel choice problem. Our work shows that it is possible to overcome the need of (wasteful and often heuristic) data splitting when designing hypothesis tests with feasible null distribution. Since this comes without relevant increase in computational resources we expect the proposed method to replace the data splitting approach in applications that fit the framework considered in this work. Theorem 1 is also applicable beyond hypothesis testing and extends the previously known PSI framework proposed by Lee et al. [24].",doesn't mention a harmful application
+Passport-aware Normalization for Deep Model Protection,https://proceedings.neurips.cc/paper/2020/file/ff1418e8cc993fe8abcfe3ce2003e5c5-Paper.pdf,"Though deep learning evolves very fast in these years, IP protection for deep models is seriously under-researched. In this work, we mainly aim to propose a general technique for deep model IP protection. It will help both academia and industry to protect their interests from illegal distribution or usage. We hope it can inspire more works along this important direction.",doesn't mention a harmful application
+Group Knowledge Transfer: Federated Learning of Large CNNs at the Edge,https://proceedings.neurips.cc/paper/2020/file/a1d4c20b182ad7137ab3606f0e3fc8a4-Paper.pdf,"FedGKT can efficiently train large deep neural networks (CNNs) in resource-constrained edge devices (such as smartphones, IoT devices, and edge servers). Unlike past FL approaches, FedGKT demonstrates the feasibility of training a large server-side model by using many small client models. FedGKT preserves the data privacy requirements of the FL approach but also works within the constraints of an edge computing environment. Smartphone users may benefit from this technique because their private data is protected, and they may also simultaneously obtain a high-quality model service. Organizations such as hospitals, and other non-profit entities with limited training resources, can collaboratively train a large CNN model without revealing their datasets while achieving significant training cost savings. They can also meet requirements regarding the protection of intellectual property, confidentiality, regulatory restrictions, and legal constraints. As for the potential risks of our method, a client can maliciously send incorrect hidden feature maps and soft labels to the server, which may potentially impact the overall model accuracy. These effects must be detected and addressed to maintain overall system stability. Second, the relative benefits for each client may vary. For instance, in terms of fairness, edge nodes which have smaller datasets may obtain more model accuracy improvement from collaborative training than those which have a larger amount of training data. Our training framework does not consider how to balance this interest of different parties.",mentions a harmful application
+Improving Local Identifiability in Probabilistic Box Embeddings,https://proceedings.neurips.cc/paper/2020/file/01c9d2c5b3ff5cbba349ec39a570b5e3-Paper.pdf,This work does not present any foreseeable societal consequence.,doesn't mention a harmful application
+A Finite-Time Analysis of Two Time-Scale Actor-Critic Methods,https://proceedings.neurips.cc/paper/2020/file/cc9b3c69b56df284846bf2432f1cba90-Paper.pdf,"This work could positively impact the industrial application of actor-critic algorithms and other reinforcement learning algorithms. The theorem exhibits the sample complexity of actor-critic algorithms, which could be used to estimate required training time of reinforcement learning models. Another direct application of our result is to set the learning rate according to the finite-time bound, by optimizing the constant factors of the dominant terms. In this sense, the result could potentially reduce the overhead of hyper-parameter tuning, thus saving both human and computational resources. Moreover, the new analysis in this paper can potentially help people in different fields to understand the broader class of two-time scale algorithms, in addition to actor-critic methods. To our knowledge, this algorithm and theory studied in our paper do not have any ethical issues.",doesn't mention a harmful application
+Active Invariant Causal Prediction: Experiment Selection through Stability,https://proceedings.neurips.cc/paper/2020/file/b197ffdef2ddc3308584dce7afa3661b-Paper.pdf,"Any method that learns from finite data is subject to statistical estimation errors and model assumptions that necessarily limit the full applicability of its findings. Unfortunately, study outcomes are not always communicated with the required qualifications. As an example, statistical hypothesis testing is often employed carelessly, e.g. by using p-values to claim “statistical significance” without paying attention to the underlying assumptions [5]. There is a danger that this problem gets exacerbated when one aims to estimate causal structures. Estimates from causal inference algorithms could be claimed to “prove” a given causal relationship, ruling out various alternative explanations that one would consider when explaining a statistical association. For example, ethnicity could be claimed to have a causal effect on criminality and thereby used as a justification for oppressive political measures. While this would represent a clear abuse of the technology, we as researchers have to ensure that similar mistakes in interpretation are not made unintentionally. This implies being conscientious about understanding as well as stating the limitations of our research. While there is a risk that causal inference methods are misused as described above, there is of course also an array of settings where causal learning—and in particular active causal learning—can be extremely useful. As our main motivation we envision the empirical sciences where interventions correspond to physical experiments which can be extremely costly in terms of time and/or money. For complex systems, as for example gene regulatory networks in biology, it might be difficult for human scientists to choose informative experiments, particularly if they are forced to rely on data alone. Our goal is to develop methods to aid scientists to better understand their data and perform more effective experiments, resulting in significant resource savings. The specific impact of our proposed methodology will depend on the application. For the method we propose in this work, one requirement for application would be that the experiments yield more than one data point (and ideally many), so that our invariance-based approach can be employed. In future work, we aim to develop methodology that is geared towards the setting where only very few data points per experiment are available.",doesn't mention a harmful application
+Continuous Meta-Learning without Tasks,https://proceedings.neurips.cc/paper/2020/file/cc3f5463bc4d26bc38eadc8bcffbc654-Paper.pdf,"Our work provides a method to extend meta-learning algorithms beyond the task-segmented case, to the time series series domain. Equivalently, our work extends core methods in changepoint detection, enabling the use of highly expressive predictive models via empirical Bayes. This work has the potential to extend the domain of applicability of both of these methods. Standard meta-learning relies on a collection of datasets, each corresponding to discrete tasks. A natural question is how such datasets are constructed; in many cases, these datasets rely on segmentation of time series data by experts. Thus, our work has the potential to make meta-learning algorithms applicable to problems that, previously, would have been too expensive or impossible to segment. Moreover, our work has the potential to improve the applicability of changepoint detection methods to difficult time series forecasting problems. While MOCA has the potential to expand the domain of problems addressable via meta-learning, this has the effect of amplifying the risks associated with these methods. Meta-learning enables efficient learning for individual members of a population via leveraging empirical priors. There are clear risks in few-shot learning generally: for example, efficient facial recognition from a handful of images has clear negative implications for privacy. Moreover, while there is promising initial work on fairness for meta-learning [39], we believe considerable future research is required to understand the degree to which meta-learning algorithms increase undesirable bias or decrease fairness. While it is plausible that fine-tuning to the individual results in reduced bias, there are potential unforeseen risks associated with the adaptation process, and future research should address how bias is potentially introduced in this process. Relative to decision making rules that are fixed across a population, algorithms which fine-tune decision making to the individual present unique challenges in analyzing fairness. Further research is required to ensure that the adaptive learning enabled by algorithms such as MOCA do not lead to unfair outcomes.",mentions a harmful application
+Learning Rich Rankings,https://proceedings.neurips.cc/paper/2020/file/6affee954d76859baa2800e1c49e2c5d-Paper.pdf,"Flexible ranking distributions that can be learned with provable guarantees can facilitate more powerful and reliable ranking algorithms inside recommender systems, search engines, and other ranking-based technological products. As a potential adverse consequence, more powerful and reliable learning algorithms can lead to an increased inappropriate reliance on technological solutions to complex problems, where practitioners may be not fully grasp the limitations of our work, e.g. independence assumptions, or that our risk bounds, as established here, do not hold for all data generating processes.",mentions a harmful application
+Reinforcement Learning for Control with Multiple Frequencies,https://proceedings.neurips.cc/paper/2020/file/216f44e2d28d4e175a194492bde9148f-Paper.pdf,"In recent years, reinforcement learning (RL) has shown remarkable successes in various areas, where most of their results are based on the assumption that all decision variables are simultaneously determined at every discrete time step. However, many real-world sequential decision-making problems involve multiple decision variables whose control frequencies are different by the domain requirement. In this situation, standard RL algorithms without considering the control frequency requirement may suffer from severe performance degradation as discussed in Section 3. This paper provides a theoretical and algorithmic foundation of how to address multiple control frequencies in RL, which enables RL to be applied to more complex and diverse real-world problems that involve decision variables with different frequencies. Therefore, this work would be beneficial for those who want to apply RL to various tasks that inherently have multiple control frequencies. As we provide a general-purpose methodology, we believe this work has little to do with a particular system failure or a particular data bias. On the other hand, this work could contribute to accelerating industrial adoption of RL, which has the potential to adversely affect employment due to automation.",mentions a harmful application
+Latent Dynamic Factor Analysis of High-Dimensional Neural Recordings,https://proceedings.neurips.cc/paper/2020/file/beb04c41b45927cf7e9f8fd4bb519e86-Paper.pdf,"While progress in understanding the brain is improving life through research, especially in mental health and addiction, in no case is any brain disorder well understood mechanistically. Faced with the reality that each promising discovery inevitably reveals new subtleties, one reasonable goal is to be able to change behavior in desirable ways by modifying specific brain circuits and, in animals, technologies exist for circuit disruptions that are precise in both space and time. However, to determine the best location and time for such disruptions to occur, with minimal off-target effects, will require far greater knowledge of circuits than currently exists: we need good characterizations of interactions among brain regions, including their timing relative to behavior. The over-arching aim of our research is to provide methods for describing the flow of information, based on evolving neural activity, among multiple regions of the brain during behavioral tasks. Such methods can lead to major advances in experimental design and, ultimately, to far better treatments than currently exist.",doesn't mention a harmful application
+Reducing Adversarially Robust Learning to Non-Robust PAC Learning,https://proceedings.neurips.cc/paper/2020/file/a822554e5403b1d370db84cfbc530503-Paper.pdf,"Learning predictors that are robust to adversarial perturbations is an important challenge in contem- porary machine learning. Current machine learning systems have been shown to be brittle against different notions of robustness such as adversarial perturbations [Szegedy et al., 2013, Biggio et al., 2013, Goodfellow et al., 2014], and there is an ongoing effort to devise methods for learning predictors that are adversarially robust. As machine learning systems become increasingly integrated into our everyday lives, it becomes crucial to provide guarantees about their performance, even when they are used outside their intended conditions. We already have many tools developed for standard learning, and having a universal wrapper that can take any standard learning method and turn it into a robust learning method could greatly simplify the development and deployment of learning that is robust to test-time adversarial perturbations. The results that we present in this paper are still mostly theoretical, and limited to the realizable setting, but we expect and hope they will lead to further theoretical study as well as practical methodological development with direct impact on applications. In this work we do not deal with training-time adversarial attacks, which is a major, though very different, concern in many cases. As with any technology, having a more robust technology can have positive and negative societal consequences, and this depends mainly on how such technology is utilized. Our intent from this research is to help with the design of robust machine learning systems for application domains such as healthcare and transportation where its critical to ensure performance guarantees even outside intended conditions. In situations where there is a tradeoff between robustness and accuracy, this work might be harmful in that it would prioritize robustness over accuracy and this may not be ideal in some application domains.",mentions a harmful application
+Online Non-Convex Optimization with Imperfect Feedback,https://proceedings.neurips.cc/paper/2020/file/c7c46d4baf816bfb07c7f3bf96d88544-Paper.pdf,This is a theoretical work which does not present any foreseeable societal consequence.,doesn't mention a harmful application
+Digraph Inception Convolutional Networks,https://proceedings.neurips.cc/paper/2020/file/cffb6e2288a630c2a787a64ccc67097c-Paper.pdf,"GCNs could be applied to a wide range of applications, including image segmentation [27], speech recognition [14], recommender system [17], point cloud [50, 24], traffic prediction [25] and many more [45]. Our method can help to expand the graph types from undirected to directed in the above application scenarios and obtain multi-scale features from the high-order hidden directed structure. For traffic prediction, our method can be used in map applications to obtain more fine-grained and accurate predictions. This requires users to provide location information, which has a risk of privacy leakage. The same concerns also arise in social network analysis [38], person re-ID [35] and NLP [49], which use graph convolutional networks as their feature extraction methods. Another potential risk is that our model may be adversarial attacked by adding new nodes or deleting existing edges. For example, in a graph-based recommender system, our model may produce completely different recommendation results due to being attacked. We see opportunities for research applying DiGCN to beneficial purposes, such as investigating the ability of DiGCN to discover hidden complex directed structure, the limitation of approximate method based on personalized PageRank and the feature oversmoothing problem in digraphs. We also encourage follow-up research to design derivative methods for different tasks based on our method.",mentions a harmful application
+Learning Physical Constraints with Neural Projections,https://proceedings.neurips.cc/paper/2020/file/37bc5e7fb6931a50b3464ec66179085f-Paper.pdf,This research constitutes a technical advance by employing constraint projection operations to enhance the prediction capability of physical systems with unknown dynamics. It opens up new possibilities to effectively and intuitively represent complicated physical systems from direct and limited observation. This research blend the borders among the communities of machine learning and fast physics simulations in computer graphics and gaming industry. Our model does not necessarily bring about any significant ethical considerations.,doesn't mention a harmful application
+Sub-sampling for Efficient Non-Parametric Bandit Exploration,https://proceedings.neurips.cc/paper/2020/file/3ab6be46e1d6b21d59a3c3a0b9d0f6ef-Paper.pdf,"This work is advertising a new way to do non-parametric exploration in bandit models, that enjoy good empirical performance and strong theoretical guarantees. First, bandit problems are at the heart of numerous applications to online content recommendation, hence the good performance of SDA algorithms may inspire new algorithms for more realistic models used for these applications, such as contextual bandits. Then, exploration is a central question in the broader field of reinforcement learning, hence new ideas for bandits may lead to new ideas for reinforcement learning.",doesn't mention a harmful application
+The Discrete Gaussian for Differential Privacy,https://proceedings.neurips.cc/paper/2020/file/b53b3a3d6ab90ce0268229151c9bde11-Paper.pdf,"We have provided a thorough analysis of the privacy and utility properties of the discrete Gaussian and the practicality of sampling it. The impact of this work is that it makes the real-world deployment of differential privacy more practical and secure. In particular, we bridge the gap between the theory (which considers continuous distributions) and the practice (where precision is finite and numerical errors can cause a dramatic privacy failures). We hope that the discrete Gaussian will be used in practice and, further, that our work is critical to enabling these real-world deployments. The positive impact of this work is clear: Differential privacy provides a principled and quantitative way to balance rigorous privacy guarantees and statistical utility in data analysis. If this technology is adopted, it can provide untrusted third parties controlled access to data (e.g., to enable scientific research), while affording the data subjects (i.e., the general public) an adequate level of privacy protection. In any case, our methods are better than using flawed methods (i.e., naïve floating-point implementations of continuous distributions) that inject noise without actually protecting privacy or using methods (such as rounding or discrete Laplace) that offer a worse privacy-utility tradeoff. The negative impact of this work is less clear. All technologies can be misused. For example, an organization may be able to deceptively claim that their system protects privacy on the basis that it is differentially private, when, in reality, it is not private at all, because their privacy parameter is enormous (e.g., ε = 10 6 ). One needs to be careful and critical about promises made by such companies, and educate the general audience about what differential privacy does provide, what it does not, and when guarantees end up being meaningless. However, we must acknowledge that there is a small – but vocal – group of people who do not want differential privacy to be deployed in practice. In particular, the US Census Bureau’s planned adoption of differential privacy for the 2020 US Census has met staunch opposition from some social scientists. We cannot speak for the opponents of differential privacy; many of their objections do not make sense to us and thus it would be inappropriate for us to try summarizing them. However, there is a salient point that needs to be discussed: Differential privacy provides a principled and quantitative way to balance rigorous privacy guarantees and statistical utility in data analysis. This is good, in theory, but, in practice, privacy versus utility is a heated and muddy debate. On one hand, data users (such as social scientists) want unfettered access to the raw data. On the other hand, privacy advocates want the data locked up or never collected in the first place. The technology of differential privacy offers a vehicle for compromise. Yet, some parties are not interested in compromise. In particular, users of census data users are accustomed to largely unrestricted data access. From a privacy perspective, this is unsustainable – the development of reconstruction attacks and the availability of large auxiliary datasets for linking/re-identification has shown that census data needs more robust protections. Understandably, those who rely on census data are deeply concerned about anything that may compromise their ability to conduct research. The adoption of differential privacy has prompted uncomfortable (but necessary) discussions about the value of providing data access relative to the privacy cost. In particular, it is necessary to decide how to allocate the privacy budget – which statistics are most important to release accurately? Another dimension of the privacy-versus-utility debate is how it affects small communities, such as racial/ethnic minorities or rural populations. Smaller populations inherently suffer a harsher privacy- utility tradeoff. Differential privacy is almost always defined so that it provides every person with an equal level of privacy. Consequently, differentially private statistics for smaller populations (e.g., Native Americans in a small settlement) will be less accurate than for larger populations (e.g., Whites in a large US city). More precisely, noise addition methods like ours offer the same absolute accuracy on all populations, but the relative accuracy will be worse when the denominator (i.e., population size) is smaller. The only alternative is to offer small communities weaker privacy protections. We stress that this issue is not specific to differential privacy. For example, if we rely on anonymity or de-identification, then we must grapple with the fact that minorities are more easily re-identified, since, by definition, minorities are more unique. This is a fundamental tradeoff that needs to be carefully considered with input from the minorities and communities concerned. Ultimately, computer scientists can only provide tools and it is up to policymakers in government and other organizations to decide how to use them. This work, along with the broader literature on differential privacy, provides such tools. However, the research community also has a responsibility to provide instructions for how these tools should and should not be used.",mentions a harmful application

data/semiconductor_org_types/task.json

@@ -0,0 +1 @@
+{"name": "semiconductor_org_types", "description": "", "data_columns": ["Paper title", "Organization name"], "label_columns": {"Label": ["company", "research institute", "university"]}}

data/semiconductor_org_types/test_unlabeled.csv

@@ -0,0 +1,450 @@
+Paper title,Organization name
+An enhanced 130 nm generation logic technology featuring 60 nm transistors optimized for high performance and low power at 0.7 - 1.4 V,"Portland Technology Development, Hillsboro, OR, USA"
+Monolithic integration of O-band photonic transceivers in a “zero-change” 32nm SOI CMOS,"Department of Electrical Engineering and Computer Science, University of California, Berkeley, USA"
+"High-performance low-leakage enhancement-mode high-K dielectric GaN MOSHEMTs for energy-efficient, compact voltage regulators and RF power amplifiers for low-power mobile SoCs","Intel Corporation,Components Research,Technology and Manufacturing Group,Hillsboro,OR,USA"
+"Implant-Free SiGe Quantum Well pFET: A novel, highly scalable and low thermal budget device, featuring raised source/drain and high-mobility channel","Dep. Material Engineering, Univ. Leuven, Leuven"
+Mobility in high-K metal gate UTBB-FDSOI devices: From NEGF to TCAD perspectives,"STMicroelectronics, Crolles, France"
+Fast switching and long retention Fe-O ReRAM and its switching mechanism,"Advanced Devices Development Center, Matsushita Elecrric Indusrrial Company Limited, Moriguchi, Osaka, Japan"
+5.9 An 18.75µW dynamic-distributing-bias temperature sensor with 0.87°C(3σ) untrimmed inaccuracy and 0.00946mm2 area,"TSMC,Austin,TX,United States of America"
+A 1-V 299/spl mu/W Flashing UWB Transceiver Based on Double Thresholding Scheme,"Center for collaborative Res.,Tokyo Univ."
+"High performance low temperature activated devices and optimization guidelines for 3D VLSI integration of FD, TriGate, FinFET on insulator","IMEP-LAHC,Minatec/INPG,France"
+Quantitative assessment of mobility degradation by remote Coulomb scattering in ultra-thin oxide MOSFETs: measurements and simulations,"DIEGM, Udine, Italy"
+A new cell-based performance metric for novel CMOS device architectures,"Philips Research, Leuven, Belgium"
+"In-situ multi-step (IMS) CVD process of (Ba,Sr)TiO/sub 3/ using hot wall batch type reactor for DRAM capacitor dielectrics","Microelectron. Eng. Lab.,Toshiba Corp.,Yokohama,Japan"
+A novel method for evaluating electron/hole mismatch in scaled split-gate SONOS memories,"Microcomputer Operation Unit, NECEL Corporation, Sagamihara, Kanagawa, Japan"
+A 120mm2 16Gb 4-MLC NAND Flash Memory with 43nm CMOS Technology,"SanDisk,Yokohama,Japan"
+High power 4H-SiC static induction transistors,"Westinghouse Science and Technology Center, Pittsburgh, PA, USA"
+Nano-wires for room temperature operated hybrid CMOS-NANO integrated circuits,"Swiss Fed. Inst. of Technol.,Lausanne,Switzerland"
+Clock-powered CMOS VLSI graphics processor for embedded display controller application,"Synopsys Corporation,Mountain View,CA,USA"
+Thermally robust high quality HfN/HfO/sub 2/ gate stack for advanced CMOS devices,"Institute of Microelectronics, Singapore"
+A 622 Mb/s fully-integrated optical IC with a wide range input,"Sony Corp.,Kanagawa,Japan"
+Fabrication of a nonvolatile lookup-table circuit chip using magneto/semiconductor-hybrid structure for an immediate-power-up field programmable gate array,"Hitachi Advanced Research Laboratory,Tokyo,,Japan"
+The impact of sub monolayers of HfO/sub 2/ on the device performance of high-k based transistors [MOSFETs],"Renesas, Leuven, Belgium"
+30.5 A 0.5V BLE Transceiver with a 1.9mW RX Achieving −96.4dBm Sensitivity and 4.1dB Adjacent Channel Rejection at 1MHz Offset in 22nm FDSOI,"Sony LSI Design,Atsugi,Japan"
+Optimization of Sub-Melt Laser Anneal: Performance and Reliability,"K. U. Leuven, ESAT-INSYS, Belgium"
+A 0.5-28GB/S Wireline Tranceiver with 15-Tap DFE and Fast-Locking Digital CDR in 7NM FinFET,"Xilinx,Inc.,San Jose,CA,USA"
+Dual channel FinFETs as a single high-k/metal gate solution beyond 22nm node,"Intel Assignee, USA"
+50 nm-Gate All Around (GAA)-Silicon On Nothing (SON)-devices: a simple way to co-integration of GAA transistors within bulk MOSFET process,"R&D France Telecom,Grenoble,France"
+Temperature compensation of silicon micromechanical resonators via degenerate doping,"School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA"
+A phase change memory cell with metallic surfactant layer as a resistance drift stabilizer,"ULVAC, Inc.,14 Suyama Susono, Shizuoka, Japan"
+"A flexible, lightweight Braille sheet display with plastic actuators driven by an organic field-effect transistor active matrix","National Institute for Advanced Industrial Science and Technology, Osaka, Japan"
+Oxide-field dependence of the NMOS hot-carrier degradation rate and its impact on AC-lifetime prediction,"Department of Electrical Engineering & Computer Science, Microsystems Technology Laboratories, Massachusetts Institute of Technology, Cambridge, MA, USA"
+Experimental observation and physics of “negative” capacitance and steeper than 40mV/decade subthreshold swing in Al0.83In0.17N/AlN/GaN MOS-HEMT on SiC substrate,"Components Research, Intel Corporation, Hillsboro, USA"
+Application-oriented performance of RF CMOS technologies on flexible substrates,"Institut d'Electronique de Microélectronique et de Nanotechnologie - IEMN UMR8520, Grenoble, France"
+A novel SPRAM (SPin-transfer torque RAM)-based reconfigurable logic block for 3D-stacked reconfigurable spin processor,"Advanced Research Laboratory, Hitachi and Limited, Sendai, Japan"
+Asymmetrically-dope buried layer (ADB) structure CMOS for low-voltage mixed analog-digital applications,"Semicond. Dev. Center,Hitachi Ltd.,Kokubunji,Japan"
+Highly sensitive and reliable X-ray detector with HgI2 photoconductor and oxide drive TFT,"Samsung Advanced Institute of Technology, Samsung Electronics Corporation, South Korea"
+Large scale plane-wave based density-functional theory simulations for electronic devices,"Lawrence Berkeley National Laboratory,Berkeley,CA,USA"
+A 76dBΩ 1.7GHz 0.18µm CMOS tunable transimpedance amplifier using broadband current pre-amplifier for high frequency lateral micromechanical oscillators,"Georgia Institute of Technology,Atlanta,USA"
+Coulomb oscillations in 100 nm and 50 nm CMOS devices,"Departement de Recherche Fondamentale sur la Matiere Condensee, DSM, Grenoble, France"
+Oxide thin film transistor technology: Capturing device-circuit interactions,"IGNIS Innovation Inc., Waterloo, ON, Canada"
+Tetragonal Phase Stabilization by Doping as an Enabler of Thermally Stable HfO2 based MIM and MIS Capacitors for sub 50nm Deep Trench DRAM,"Qimonda Dresden GmbH and Company OHG, Dresden, Germany"
+Light emitting silicon nanostructures,"Charles Stark Draper Laboratories, Inc., Cambridge, MA, USA"
+Effective Schottky Barrier Height modulation using dielectric dipoles for source/drain specific contact resistivity improvement,"College of Nanoscale Science and Engineering, Albany, NY, USA"
+Energy-efficient all fiber-based local body heat mapping circuitry combining thermistor and memristor for wearable healthcare device,"KIMS Changwon, Korea"
+"Bidirectional TaO-diode-selected, complementary atom switch (DCAS) for area-efficient, nonvolatile crossbar switch block","Low-power Electronics Association & Project (LEAP),West,Onogawa,Tsukuba,Ibaraki,Japan"
+A 1/2.5 inch 8.1Mpixel CMOS Image Sensor for Digital Cameras,"Micron Technology,Pasadena,CA"
+A 0.2-/spl mu/m 180-GHz-f/sub max/ 6.7-ps-ECL SOI/HRS self aligned SEG SiGe HBT/CMOS technology for microwave and high-speed digital applications,"Musashino office, Hitachi Device Engineering Company Limited, Japan"
+A unified physical model of switching behavior in oxide-based RRAM,"NASA Ames Research Center,Moffett Field,CA,USA"
+I.McIC: A single-chip MPEG2 video encoder for storage,"Philips Res. Lab.,Eindhoven,Netherlands"
+A highly linear filter and VGA chain with novel DC-offset correction in 90nm digital CMOS process,"Intel R&D,Intel Corp.,Hillsboro,OR,USA"
+A high performance phase change memory with fast switching speed and high temperature retention by engineering the GexSbyTez phase change material,"Macronix Emerging Central Laboratory, Macronix International Company Limited, Hsinchu, Taiwan"
+Adaptive cancellation of gain and nonlinearity errors in pipelined ADCs,"Asahi Kasei Microdevices,Atsugi,Japan"
+A 4 GOPS 3 way-VLIW image recognition processor based on a configurable media-processor,"Toshiba Corp.,Kanagawa,Japan"
+Hot carrier reliability for 0.13 /spl mu/m CMOS technology with dual gate oxide thickness,"UMC, Hopewell Junction, NY, USA"
+Implementation of the CELL Broadband Engine in a 65nm SOI Technology Featuring Dual-Supply SRAM Arrays Supporting 6GHz at 1.3V,"Toshiba,Austin,TX"
+A fully working 0.14 /spl mu/m DRAM technology with polymetal (W/WN/sub x//poly-Si) gate,"Hyundai Electron. Ind. Co. Ltd., Cheongju, South Korea"
+Potential well engineering by partial oxidation of TiN for high-speed and low-voltage Flash memory with good 125°C data retention and excellent endurance,"Thin-Film Materials Research Center, Korea Institute of Science and Technology, Seoul, South Korea"
+A programmable MEMS-based clock generator with sub-ps jitter performance,"Masdar Institute,Abu Dhabi,UAE"
+Impact of Fermi level pinning inside conduction band on electron mobility of InxGa1−xAs MOSFETs and mobility enhancement by pinning modulation,"National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan"
+Equivalent Oxide Thickness (EOT) Scaling With Hafnium Zirconium Oxide High-κ Dielectric Near Morphotropic Phase Boundary,"Kurt J. Lesker Co., PA, USA"
+3.3 A 5GS/s 158.6mW 12b Passive-Sampling 8×-Interleaved Hybrid ADC with 9.4 ENOB and 160.5dB FoMS in 28nm CMOS,"KU Leuven,Heverlee,Belgium"
+Characteristics of AlGaN/GaN HEMT devices with SiN passivation,"DaimlerChrysler AG Research and Technology, Ulm, Germany"
+A novel resist and post-etch residue removal process using ozonated chemistries,"IMEC,Philips Research,Eindhoven,Netherlands"
+14.5 Envision: A 0.26-to-10TOPS/W subword-parallel dynamic-voltage-accuracy-frequency-scalable Convolutional Neural Network processor in 28nm FDSOI,"KU Leuven,Belgium"
+A 0.063 µm2 FinFET SRAM cell demonstration with conventional lithography using a novel integration scheme with aggressively scaled fin and gate pitch,"Toshiba America Electronic Components Inc.,Albany Nano Tech,NY,USA"
+Scaling rules of piezoelectric nanowires in view of sensor and energy harvester integration,"CEA-Leti, Grenoble, France"
+A 0.9 V 1.5 mW continuous-time /spl Delta//spl Sigma/ modulator for WCDMA,"Toshiba,Kawasaki,Japan"
+A study for 0.18 /spl mu/m high-density MRAM,"Technol. Dev. Group,Sony Corp.,Kanagawa,Japan"
+A Fractional-N PLL for SONET-Quality Clock-Syntlhesis Applicationis,"Silicon Laboratories,Nashua,NH"
+On-line calibration and digital correction of multi-bit sigma-delta modulators,"Dept. of Electr. Eng.,Pavia Univ.,Italy"
+Intrinsic retention statistics in phase change memory (PCM) arrays,"Micron, R&D Unit, Agrate Brianza, Italy"
+Multi-level metal CMOS manufacturing with deuterium for improved hot carrier reliability,"Lucent Technologies Bell Laboratories, Orlando, FL, USA"
+An adaptive reference generation scheme for 1T1C FeRAMs,"Dept. of Electr. & Comput. Eng.,Toronto Univ.,Ont.,Canada"
+Ferroelectric hafnium oxide: A CMOS-compatible and highly scalable approach to future ferroelectric memories,"Fraunhofer IPMS-CNT, Dresden, Germany"
+High-Field Electron Mobility in Biaxially-tensile Strained SOI: Low Temperature Measurement and Correlation with the Surface Morphology,"CEA/LETI Minatec,rue des Martyrs,Grenoble,France"
+Threshold voltage control in NiSi-gated MOSFETs through silicidation induced impurity segregation (SIIS),"Microelectronics Division, Hopewell Junction, NY, USA"
+Higher hole mobility induced by twisted Direct Silicon Bonding (DSB),"IBM Research Division,T.J. Watson Research Center,Yorktown Heights,NY,USA."
+Realizing a production ATE custom processor and timing IC containing 400 independent low-power and high-linearity timing verniers,"Credence Syst.,Fremont,CA,USA"
+22.4 A 24Gb/s 0.71pJ/b Si-photonic source-synchronous receiver with adaptive equalization and microring wavelength stabilization,"Hewlett-Packard Labs,Palo Alto,CA"
+A 7.9μW remotely powered addressed sensor node using EPC HF and UHF RFID technology with −10.3dBm sensitivity,"Infineon Technologies,Graz,Austria"
+Bridging design and manufacture of analog/mixed-signal circuits in advanced CMOS,"AMD,Inc.,Sunnyvale,E. Arques Ave.,CA,USA"
+Experience of IP-reuse in system-on-chip design for ADSL,"Alcatel Bell Telephone,Antwerp,Belgium"
+Analysis of trap-assisted conduction mechanisms through silicon dioxide films using quantum yield,"Bell Laboratories, Lucent Technologies, Inc., Murray Hill, NJ, USA"
+SRAM Cell Static Noise Margin and VMIN Sensitivity to Transistor Degradation,"Silicon Technology Development, Dallas, TX, USA"
+"Vacancy-modulated conductive oxide resistive RAM (VMCO-RRAM): An area-scalable switching current, self-compliant, highly nonlinear and wide on/off-window resistive switching cell","K.U. Leuven, Leuven, Belgium"
+Highly Reliable Thin MIM Capacitor on Metal (CoM) Structure with Vertical Scalability for Analog/RF Applications,"NEC Corporation Limited, Sagamihara, Kanagawa, Japan"
+A 4.4mW 76dB complex /spl Sigma//spl Delta/ ADC for Bluetooth receivers,"Philips Res. Labs.,Eindhoven,Netherlands"
+"A 210mV 7.3MHz 8T SRAM with dual data-aware write-assists and negative read wordline for high cell-stability, speed and area-efficiency","Fukuoka Institute of Technology,Japan"
+Digital background calibration of a 10 b 40 M sample/s parallel pipelined ADC,"California Univ.,Davis,CA,USA"
+Aluminum Plasma-CVD for VLSI Circuit Interconnections,"Fujitsu Laboratories Ltd. Kamikodanaka,Nakahara,Kawasaki,Japan"
+A 1.4GHz 20.5Gbps GZIP decompression accelerator in 14nm CMOS featuring dual-path out-of-order speculative Huffman decoder and multi-write enabled register file array,"Circuits Research Lab,Intel Corporation,Hillsboro,OR,USA"
+4.4 Energy-efficient microserver based on a 12-core 1.8GHz 188K-CoreMark 28nm bulk CMOS 64b SoC for big-data applications with 159GB/S/L memory bandwidth system density,"Freescale Semiconductor,Austin,TX"
+Understanding and Physical Modeling Superior Hot-Carrier Reliability of Ge pNWFETs,"Nanolayers, London, UK"
+Lattice strain design in W/WN/poly-Si gate DRAM for improving data retention time,"System Devices Research Laboratories, NEC Corporation Limited, Sagamihara, Kanagawa, Japan"
+Highly endurable floating body cell memory: Vertical biristor,"Department of EE, KAIST, Daejeon, South Korea"
+A 58.6mW real-time programmable object detector with multi-scale multi-object support using deformable parts model on 1920×1080 video at 30fps,"Massachusetts Institute of Technology,USA"
+"Multipurpose, Fully-Integrated 128×128 Event-Driven MD-SiPM with 512 16-Bit TDCs with 45 PS LSB and 20 NS Gating","EPFL,Switzerland"
+A 5Gb/s link with clock edge matching and embedded common mode clock for low power interfaces,"NVIDIA Corporation,India"
+A new vertically stacked poly-Si MOSFET for 533 MHz high speed 64Mbit SRAM,"Renesas Technology Corp., Tokyo, Japan"
+The implementation of POWER7TM: A highly parallel and scalable multi-core high-end server processor,"IBM,Poughkeepsie,NY,USA"
+A Machine-Learning-Resistant 3D PUF with 8-layer Stacking Vertical RRAM and 0.014% Bit Error Rate Using In-Cell Stabilization Scheme for IoT Security Applications,"Zhejiang Lab,Hangzhou,China"
+45nm low power CMOS logic compatible embedded STT MRAM utilizing a reverse-connection 1T/1MTJ cell,"Qualcomm Incorporated, San Diego, CA, USA"
+Full integration and characterization of Localized ONO Memory (LONOM) for embedded flash technology,"Syst. LSI Div.,Samsung Electron. Co. Ltd.,Kyunggi-Do,South Korea"
+A digital terrestrial television (ISDB-T) tuner for mobile applications,"Sharp Corp.,Tenri,Japan"
+The roles of hydrogen and holes in trap generation and breakdown in ultra-thin SiON dielectrics,"Silicon Technology Development, Texas Instruments, Inc., Dallas, TX, USA"
+Memory technology for the terabit era: From 2D to 3D,"KU Leuven,ESAT Department,Leuven,Belgium,and imec,Leuven,Belgium"
+"A 1.5 V, 4.1 mW dual channel audio delta-sigma D/A converter","Asahi-Kasei Microsyst.,Kanagawa,Japan"
+CMOS current-controlled oscillators using multiple-feedback-loop ring architectures,"Korea Adv. Energy Res. Inst.,Taejon,South Korea"
+Aggressive design of millisecond annealing junctions for near-scaling-limit bulk CMOS using raised source/drain extensions,"NEC Informatec Systems Limited, Sagamihara, Japan"
+Simultaneous Extraction of Recoverable and Permanent Components Contributing to Bias-Temperature Instability,"IMEC, Leuven, Belgium"
+A Fully Digital 65nm CMOS Transmitter for the 2.4-to-2.7GHz WiFi/WiMAX Bands using 5.4GHz ΔΣ RF DACs,"STMicroelectronics,Geneva,Switzerland"
+Low RA Magnetic Tunnel Junction Arrays in Conjunction with Low Switching Current and High Breakdown Voltage for STT-MRAM at 10 nm and Beyond,"Corporate Research and Development,Qualcomm Technologies,Inc.,San Diego,CA,USA"
+A CMOS 6b 400 M sample/s ADC with error correction,"Fujitsu VLSI Limited,Aichi,Japan"
+A digital wideband CDR with ±15.6kppm frequency tracking at 8Gb/s in 40nm CMOS,"Broadcom,Irvine,CA"
+Understanding of Tunable Selector Performance in Si-Ge-As-Se OTS Devices by Extended Percolation Cluster Model Considering Operation Scheme and Material Design,"IMEC,Leuven,Belgium"
+Self-limiting laser thermal process for ultra-shallow junction formation of 50-nm gate CMOS,"Device Development Center, Hitachi and Limited, Ome, Tokyo, Japan"
+Soft error considerations for deep-submicron CMOS circuit applications,"Intel Corporation, Hudson, MA, USA"
+A Low Power Continuous-Time Zoom ADC for Audio Applications,"NXP Semiconductors,Eindhoven,The Netherlands"
+Dual-damascene interconnects with 0.28 /spl mu/m vias using in situ copper doped aluminum chemical vapor deposition,"ULSI Device Develop. Laboratories, NEC Corporation Limited, Sagamihara, Kanagawa, Japan"
+"Experimental study on BTI variation impacts in SRAM based on high-k/metal gate FinFET: From transistor level Vth mismatch, cell level SNM to product level Vmin","Quality and Reliability Team, Samsung Electronics Co. Ltd., Yongin-City, Gyeonggi-Do, Korea"
+Enabling Efficient Design-Technology Interaction by Spec-Driven Extraction Flow,"ProPlus Design Solutions,Inc,San Jose,CA,USA"
+Understanding of Tunable Selector Performance in Si-Ge-As-Se OTS Devices by Extended Percolation Cluster Model Considering Operation Scheme and Material Design,"Applied Materials Inc.,Santa Clara,CA,USA"
+Redefinition of Write Margin for Next-Generation SRAM and Write-Margin Monitoring Circuit,"NEC,Sagamihara"
+High on/off-ratio P-type oxide-based transistors integrated onto Cu-interconnects for on-chip high/low voltage-bridging BEOL-CMOS I/Os,"LSI Research Laboratory, Renesas Electronics Corporation, Sagamihara, Kanagawa, Japan"
+29.5 A Single-Chip Optical Phased Array in a 3D-Integrated Silicon Photonics/65nm CMOS Technology,"Colleges of Nanoscale Science and Engineering,Albany,NY"
+Scalable quantum computing with ion-implanted dopant atoms in silicon,"UNSW, School of Electrical Engineering & Telecommunications, Sydney, Australia"
+Low-cost gate-oxide early-life failure detection in robust systems,"NEC Corporation,Japan"
+Metal-Assisted Solid-Phase Crystallization Process for Vertical Monocrystalline Si Channel in 3D Flash Memory,"Institute of Memory Technology Research & Development, Kioxia Corporation, Yokkaichi, Japan"
+"A 0.9V 66MHz access, 0.13um 8M(256K/spl times/32) local SONOS embedded flash EEPROM","Syst. LSI Div.,Samsung Electron. Co. Ltd,Yongin,South Korea"
+A 160μW 8-channel active electrode system for EEG monitoring,"Imec - Holst Centre,Eindhoven,The Netherlands"
+A mobility enhancement strategy for sub-14nm power-efficient FDSOI technologies,"CEA, MINATEC Campus, Grenoble, France"
+CMOS device optimization for mixed-signal technologies,"Philips Research Laboratories, Eindhoven, Netherlands"
+Modeling of cumulative thermo-mechanical stress (CTMS) produced by the shallow trench isolation process for 1 Gb DRAM and beyond,"CAE, Semiconductor R&D Center, Samsung Electronics Company Limited, Yongin si, Gyeonggi, South Korea"
+A 14-bit 2.5GS/s and 5GS/s RF sampling ADC with background calibration and dither,"Analog Devices,Greensboro,NC,USA"
+A highly manufacturable high density embedded SRAM technology for 90 nm CMOS,"Semiconductor Company, Toshiba Corporation, Yokohama, Japan"
+A middle-1X nm NAND flash memory cell (M1X-NAND) with highly manufacturable integration technologies,"Research and Development Division, Hynix Semiconductor Inc., Ichon, Gyeonggi, South Korea"
+Pionics: the Emerging Science and Technology of Graphene-based Nanoelectronics,"School of Physics, Georgia Institute of Technology, USA"
+A novel sub-50 nm multi-bridge-channel MOSFET (MBCFET) with extremely high performance,"R&D Center,Samsung Electron. Co.,Kyunggi-Do,South Korea"
+A 15-GHz integrated CMOS switch with 21.5-dBm IP/sub 1dB/ and 1.8-dB insertion loss,"Dept. of Electr. & Comput. Eng.,Florida Univ.,Gainesville,FL,USA"
+16.1 A 12b 18GS/s RF Sampling ADC with an Integrated Wideband Track-and-Hold Amplifier and Background Calibration,"Analog Devices,Greensboro,NC"
+A 24mW 1.25Gb/s 13k/spl Omega/ transimpedance amplifier using active compensation,"Nat. Chiao Tung Univ.,Hsinchu"
+Single silicide comprising Nickel-Dysprosium alloy for integration in p- and n-FinFETs with independent control of contact resistance by Aluminum implant,"Institute of Microelectronics,Science Park Road,Singapore"
+"A 5,sup>th-order CT/DT Multi-Mode ΔΣ Modulator","NXP Semiconductors,Zurich,Switzerland"
+A robust array architecture for a capacitorless MISS tunnel-diode memory,"Central Res. Lab.,Hitachi Ltd.,Tokyo,Japan"
+Electrical integrity of state-of-the-art 0.13 /spl mu/m SOI CMOS devices and circuits transferred for three-dimensional (3D) integrated circuit (IC) fabrication,"IBM T. J. Watson Research Center, Yorktown Heights, NY, USA"
+Sub-60 nm deeply-scaled channel length extremely-thin body InxGa1−xAs-on-insulator MOSFETs on Si with Ni-InGaAs metal S/D and MOS interface buffer engineering,"Sumitomo Chemical Co. Ltd.,Kitah ara,Tsukuba,Ibaraki,Japan"
+Implementing application specific memory,"MOSAID Technol. Inc.,Kanata,Ont.,Canada"
+"SRAM critical yield evaluation based on comprehensive physical / statistical modeling, considering anomalous non-Gaussian intrinsic transistor fluctuations","System device research laboratories,NEC corporation,NEC corporation,simokuzawa,Sagamihara,Kanagawa Japan"
+A novel nonvolatile memory with spin torque transfer magnetization switching: spin-ram,"Semiconductor Technology Development Group, Semiconductor Solution Network Company, Sony Corporation, Atsugi, Kanagawa, Japan"
+Gait identification using stochastic OXRRAM-based time sequence machine learning,"IMEC,Kapeldreef,Leuven,,Belgium"
+1.2 Gbps/pin simultaneous bidirectional transceiver logic with bit deskew technique,"Device Dev. Center,Htachi Ltd.,Tokyo,Japan"
+Experimental results on reduced harmonic distortion in circuits with correlated double sampling,"Newport Microsyst. Inc.,Irvine,CA,USA"
+110nm NROM technology for code and data flash products,"Infineon Technol. Flash,Dresden,Germany"
+A Novel Via-fuse Technology Featuring Highly Stable Blow Operation with Large On-off Ratio for 32nm Node and Beyond,"Advanced Device Development Division, NEC Electronics Corporation Limited, Sagamihara, Kanagawa, Japan"
+On-chip integrated CMOS optical microspectrometer with light-to-frequency converter and bus interface,"Delft Univ. of Technol.,Netherlands"
+Device engineering for diamond quantum sensors,"Tokyo Institute of Technology, Meguro, Tokyo, Japan"
+Dynamic-sleep transistor and body bias for active leakage power control of microprocessors,"Intel Corp.,Hillsboro,OR,USA"
+Comparison between ultra-thin ZrO/sub 2/ and ZrO/sub x/N/sub y/ gate dielectrics in TaN or poly-gated NMOSCAP and NMOSFET devices,"Microelectron. Res. Center,Texas Univ.,Austin,TX,USA"
+Development of High-Voltage Vertical GaN PN Diodes,"Naval Postgraduate School,Monterey,CA,USA"
+Generic learning of TDDB applied to RRAM for improved understanding of conduction and switching mechanism through multiple filaments,"ESAT Department, K.U. Leuven, Belgium"
+A five stage chopper stabilized instrumentation amplifier using feedforward compensation,"Crystal Semicond. Div.,Cirrus Logic Inc.,Austin,TX,USA"
+Fully depleted extremely thin SOI for mainstream 20nm low-power technology and beyond,"IBM T. J. Watson,Yorktown Heights,NY,USA"
+Fabrication and characterisation of strained Si heterojunction bipolar transistors on virtual substrates,"KTH, Sweden"
+A configurable SRAM with constant-negative-level write buffer for low-voltage operation with 0.149µm2 cell in 32nm high-k metal-gate CMOS,"Toshiba Semiconductor,Kawasaki,Japan"
+From Interconnect Materials and Processes to Chip Level Performance: Modeling and Design for Conventional and Exploratory Concepts,"Georgia Institute of Technology,Atlanta,GA,USA"
+High-mobility 0.85nm-EOT Si0.45Ge0.55-pFETs: Delivering high performance at scaled VDD,"IMEC, Belgium"
+In-depth Investigation of Hf-based High-k Dielectrics as Storage Layer of Charge-Trap NVMs,"IMEP CNRS, MINA TEC, Grenoble, France"
+A 28nm 10Mb Embedded Flash Memory for IoT Product with Ultra-Low Power Near-1V Supply Voltage and High Temperature for Grade 1 Operation,"Samsung Electronics,,Samsungjeonja-ro,Hwaseong-si,Gyeonggi-do,Republic of Korea"
+High-performance high-κ/metal gates for 45nm CMOS and beyond with gate-first processing,"Toshiba America Electronic Components Research Center,Yorktown Heights,NY,USA"
+A 35mW8 b 8.8 GS/s SAR ADC with low-power capacitive reference buffers in 32nm Digital SOI CMOS,"IBM Research - Zurich,Rueschlikon,Switzerland"
+"A novel NAND-type PHINES nitride trapping storage flash memory cell with physically 2-bits-per-cell storage, and a high programming throughput for mass storage applications","Technol. Dev. Center,Macronix Int. Co.,Lt,Hsin-Chu,Taiwan"
+17.8 A 2.6μW Monolithic CMOS Photoplethysmographic Sensor Operating with 2μW LED Power,"EPFL,Neuchâtel,Switzerland"
+A 5Gb/s NRZ transceiver with adaptive equalization for backplane transmission,"Vitesse Semicond.,Somerset,NJ,USA"
+A 1.2V 1.33Gb/s/pin 8Tb NAND flash memory multi-chip package employing F-chip for low power and high performance storage applications,"Flash Memory Design Team,Samsung Electronics,Hwasung,Gyeonggi-do,Korea"
+A hydrogen barrier interlayer dielectric with a SiO/sub 2//SiON/SiO/sub 2/ stacked film for logic-embedded FeRAMs,"System LSI Design Engineering Division, NEC Corporation Limited, Sagamihara, Kanagawa, Japan"
+Scalable 3-D vertical chain-cell-type phase-change memory with 4F2 poly-Si diodes,"Yokohama Research Laboratory,Hitachi,Ltd.,Kanagawa,JAPAN"
+Low temperature (<500/spl deg/C) SrTiO/sub 3/ capacitor process technology for embedded DRAM,"Technol. Dev. Div.,Fujitsu Ltd.,Japan"
+New physical model for ultra-scaled 3D nitride-trapping non-volatile memories,"IMEP-LAHC, MINATEC-INPG, Grenoble, France"
+90 nm generation Cu/CVD low-k (k < 2.5) interconnect technology,"Taiwan Semiconductor Manufacturing Company, Science-Based Industrial Park, Hsin-Chu, Taiwan R.O.C"
+32-bit Processor core at 5-nm technology: Analysis of transistor and interconnect impact on VLSI system performance,"ARM Inc., Austin, TX, USA"
+A low power 6-bit flash ADC with reference voltage and common-mode calibration,"Broadcom Corporation,Irvine,CA,USA"
+Integration of silicon photonics in bulk CMOS,"Micron Technology,Inc. Process R&D,Boise,ID,USA"
+A novel integration of STT-MRAM for on-chip hybrid memory by utilizing non-volatility modulation,"Semiconductor R&D Center, Samsung Electronics, Co. Ltd., Hwaseong, South Korea"
+Enhanced time delay integration imaging using embedded CCD in CMOS technology,"imec, Leuven, Belgium"
+"Strained Si1−xGex-on-insulator PMOS FinFETs with excellent sub-threshold leakage, extremely-high short-channel performance and source injection velocity for 10nm node and beyond","GLOBALFOUNDRIES,T.J. Watson Research Center,Yorktown Heights,NY,USA"
+A 7nm Leakage-Current-Supply Circuit for LDO Dropout Voltage Reduction,"Georgia Institute of Technology,Atlanta"
+Multi-layer cross-point binary oxide resistive memory (OxRRAM) for post-NAND storage application,"Process Development Team, Samsung Electronics Co., Ltd., Yongin si, South Korea"
+9.1 A 45nm CMOS RF-to-Bits LTE/WCDMA FDD/TDD 2×2 MIMO base-station transceiver SoC with 200MHz RF bandwidth,"Texas Instruments,Bangalore,India"
+A Fully-Integrated UHF Receiver with Multi-Resolution Spectrum-Sensing (MRSS) Functionality for IEEE 802.22 Cognitive-Radio Applications,"Samsung RFIC Design Center,Atlanta,GA"
+A novel sense amplifier for flexible voltage operation NAND flash memories,"ULSI Res. Labs.,Toshiba Corp.,Kawasaki,Japan"
+"A 160 mW, 80 nA standby, MPEG-4 audiovisual LSI with 16 Mb embedded DRAM and a 5 GOPS adaptive post filter","Toshiba Corp.,Kawasaki,Japan"
+First demonstration of a back-side integrated heterogeneous hybrid III-V/Si DBR lasers for Si-photonics applications,"CEA-LETI, Grenoble Cedex 9, France"
+21.3 A 200nA single-inductor dual-input-triple-output (DITO) converter with two-stage charging and process-limit cold-start voltage for photovoltaic and thermoelectric energy harvesting,"Analog Devices,San Jose,CA,United States"
+Liner-supported cylinder (LSC) technology to realize Ru/Ta/sub 2/O/sub 5//Ru capacitor for future DRAMs,"Process & Manufacturing Engineering Center, Toshiba Corporation, Yokohama, Japan"
+Epitaxial SrTiO3 on silicon with EOT of 5.4 /spl Aring/ for MOS gate dielectric applications,"Dept. of Materials Science & Engineering, Kwangiu Institute of Science & Technology, Gwangju, KOREA"
+Reliability of thin gate oxide under plasma charging caused by antenna topography-dependent electron shading effect,"Logic Device Development Laboratory, ULSI Device Developmmt Laboratories, NEC Corporation Limited, Sagamihara, Kanagawa, Japan"
+"A 78 dB dynamic range, 0.27 dB accuracy, single-stage RF-PGA using thermometer-weighted and binary-weighted transconductors for SAW-less WCDMA/LTE transmitters","Renesas Technology Corp.,Hyogo,Japan"
+Comprehensive analysis of the impact of single and arrays of through silicon vias induced stress on high-k / metal gate CMOS performance,"Panasonic, Leuven, Belgium"
+SRAM design on 65nm CMOS technology with integrated leakage reduction scheme,"Portland Technol. Dev.,Intel Corp.,Hillsboro,OR,USA"
+A CMOS DVD 4/spl times/ speed read channel programmable over 5 octaves,"Samsung Electronics Company Limited,Suwon,South Korea"
+Comprehensive understanding of conductive filament characteristics and retention properties for highly reliable ReRAM,"Automotive & Industrial Systems Company,Kotari-yakemachi,Nagaokakyo City,Kyoto,Japan"
+A fully integrated multi-band MIMO WLAN transceiver RFIC,"Carleton Univ.,Ottawa,Ont.,Canada"
+Statistical Characterization and On-Chip Measurement Methods for Local Random Variability of a Process Using Sense-Amplifier-Based Test Structure,"IBM T. J. Watson,Yorktown Heights,NY"
+Program/erase dynamics and channel conduction in nanocrystal memories,"IFN-CNR, Milano, Italy"
+Heterogeneously integrated sub-40nm low-power epi-like Ge/Si monolithic 3D-IC with stacked SiGeC ambient light harvester,"Research Center for Applied Sciences, Academia Sinica, Taipei, Taiwan"
+More-than-Universal Mobility in Double-Gate SOI p-FETs with Sub-10-nm Body Thickness -Role of Light-Hole Band and Compatibility with Uniaxial Stress Engineering,"Advanced LSI Technology Laboratory, Toshiba Corporation, Yokohama, Japan"
+1Gbit High Density Embedded STT-MRAM in 28nm FDSOI Technology,"Foundry Business, Samsung Electronics Co., Giheung, Korea"
+A 390Mb/s 3.57mm2 3GPP-LTE turbo decoder ASIC in 0.13µm CMOS,"ETH Zürich,Switzerland"
+Experimental characterization of stiction due to charging in RF MEMS,"K.U. Leuven, Belgium"
+9.7 An LTE SAW-less transmitter using 33% duty-cycle LO signals for harmonic suppression,"MediaTek,Hsinchu,Taiwan"
+CMOS Integrated DNA Microarray Based on GMR Sensors,"Stanford Genome Technology Center, Palo Alto, CA, USA"
+A 40MHz-to-1GHz fully integrated multistandard silicon tuner in 80nm CMOS,"Marvell,Santa Clara,CA"
+A 20 Mhz BiCMOS peak detect pulse qualifier and area detect servo demodulator for hard disk drive servo loop,"Silicon Syst. Inc.,San Jose,CA,USA"
+512 Mb PROM with 8 layers of antifuse/diode cells,"Matrix Semicond.,Santa Clara,CA,USA"
+Capacity optimization of emerging memory systems: A shannon-inspired approach to device characterization,"Macronix International Co., Ltd., Emerging Central Lab, Hsinchu Science Park, Taiwan"
+Barriers to the Adoption of Wide-Bandgap Semiconductors for Power Electronics,"Advanced Research Projects Agency-Energy, U.S. Department of Energy, Washington, DC"
+Analytical model of the programming characteristics of scaled MONOS memories with a variety of trap densities and a proposal of a trap-density-modulated MONS memory,"Semiconductor Network Company, Sony Corporation, Atsugi, Kanagawa, Japan"
+2D molybdenum disulfide (MoS2) transistors driving RRAMs with 1T1R configuration,"Department of Electrical Engineering, Stanford, CA, USA"
+Ultra-thin-body and BOX (UTBB) fully depleted (FD) device integration for 22nm node and beyond,"IBM,USA"
+A high reliability metal insulator metal capacitor for 0.18 /spl mu/m copper technology,"IBM Semiconductor Research and Development Center, Hopewell Junction, NY, USA"
+A process independent 800 MB/s DRAM bytewide interface featuring command interleaving and concurrent memory operation,"Rambus Inc.,Mountain View,CA,USA"
+0.7 V SRAM Technology with Stress-Enhanced Dopant Segregated Schottky (DSS) Source/Drain Transistors for 32 nm Node,"Center for Semiconductor R&D,Semiconductor Company,Toshiba Corporation,,Shinsugita-cho,Isogo-ku,Yokohama,Japan"
+A Fully Integrated SoC for GSM/GPRS in 0.13/spl mu/m CMOS,"Infineon,Munich,Germany"
+A 10Gb/s compact low-power serial I/O with DFE-IIR equalization in 65nm CMOS,"Massachusetts Institute of Technology,Cambridge,USA"
+NbO2-based low power and cost effective 1S1R switching for high density cross point ReRAM Application,"R&D Division,SK Hynix Inc.,Gyeongchung-daero Bubal-eub,Icheon-si,Gyeonggi-do,,Korea"
+Dark current reduction in very-large area CCD imagers for professional DSC applications,"DALSA Semiconductor, Eindhoven, Netherlands"
+An energy harvesting wireless sensor node for IoT systems featuring a near-threshold voltage IA-32 microcontroller in 14nm tri-gate CMOS,"Internet of Things Group,Intel Corporation,Hillsboro,OR,USA"
+AES-based cryptographic and biometric security coprocessor IC in 0.18-/spl mu/m CMOS resistant to side-channel power analysis attacks,"Dept. of Electr. Eng.,California Univ.,Los Angeles,CA,USA"
+Non-Gaussian distribution of SRAM read current and design impact to low power memory using Voltage Acceleration Method,"Qualcomm Inc,Morehouse Drive,San Diego,CA,USA"
+Modeling of ultra-low energy boron implantation in silicon,"Eaton Corporation, Beverly, MA, USA"
+A video signal processor for motion-compensated field-rate upconversion in consumer television,"Philips Consumer Electron.,Hamburg,Netherlands"
+"Slurry engineering for self-stopping, dishing free SiO/sub 2/-CMP","Semiconductor Manufacturing Engineering Center, Toshiba Corporation, Kawasaki, Japan"
+A manufacturable 25 nm planar MOSFET technology,"Philips Res.,Leuven,Belgium"
+Z-PIM: An Energy-Efficient Sparsity Aware Processing-In-Memory Architecture with Fully-Variable Weight Precision,"KAIST,Daejeon,Republic of Korea"
+5.6 A 0.13μm fully digital low-dropout regulator with adaptive control and reduced dynamic stability for ultra-wide dynamic range,"Georgia Institute of Technology,Atlanta,GA"
+Copper drift in low-K polymer dielectrics for ULSI metallization,"Center for Integrated Syst.,Stanford Univ.,CA,USA"
+100 MHz CMOS circuits using sequential laterally solidified silicon thin-film transistors on plastic,"Sarnoff Corporation, Princeton, NJ, USA"
+"High performance 5nm radius Twin Silicon Nanowire MOSFET (TSNWFET) : fabrication on bulk si wafer, characteristics, and reliability","R&D Center, Samsung Electronics Company Limited, Yongin si, Gyeonggi, South Korea"
+A configurable 5-D packet classification engine with 4Mpacket/s throughput for high-speed data networking,"Lucent Technol.,Bell Labs.,Holmdel,NJ,USA"
+"19.6 A 0.2V trifilar-coil DCO with DC-DC converter in 16nm FinFET CMOS with 188dB FOM, 1.3kHz resolution, and frequency pushing of 38MHz/V for energy harvesting applications","TSMC,Hsinchu,Taiwan"
+Sub-quarter micron CMOS process for TiN-gate MOSFETs with TiO/sub 2/ gate dielectric formed by titanium oxidation,"Adv. Products Res. & Dev. Lab.,Motorola Inc.,Austin,TX,USA"
+Understanding stress enhanced performance in Intel 90nm CMOS technology,"Technol. CAD,Intel Corp.,Hillsboro,OR,USA"
+A 0.24-/spl mu/m/sup 2/ cell process with 0.18-/spl mu/m width isolation and 3-D interpoly dielectric films for 1-Gb flash memories,"Hitachi ULSI Engineering Corporation, Kodaira, Tokyo, Japan"
+A 2 GHz 60 dB dynamic-range Si logarithmic/limiting amplifier with low phase deviations,"NTT Syst. Electron. Labs.,Atsugi,Japan"
+Pionics: the Emerging Science and Technology of Graphene-based Nanoelectronics,"ECE, Georgia Institute of Technology, USA"
+A 1.8 V 2 Gb NAND flash memory for mass storage applications,"Samsung Electron.,Hwasung,South Korea"
+Efficiency of mechanical stressors in Planar FDSOI n and p MOSFETs down to 14nm gate length,"IMEP-LAHC,MINATEC Campus,Grenoble,France"
+23.9 An 8-channel 4.5Gb 180GB/s 18ns-row-latency RAM for the last level cache,"Piecemakers Technology,Hsinchu,Taiwan"
+A sub-nanosecond 0.5 /spl mu/m 64 b adder design,"Hewlett-Packard Co.,Fort Collins,CO,USA"
+Overcoming interconnect scaling challenges using novel process and design solutions to improve both high-speed and low-power computing modes,"Microarchitecture Research Laboratory, Intel Corporation, USA"
+A 21-channel 8Gb/s transceiver macro with 3.6ns latency in 90nm CMOS for 80cm backplane communication,"Hitachi ULSI Systems,Co.,Ltd.,Tokyo,Japan"
+Systematic optimization of 1 Gbit perpendicular magnetic tunnel junction arrays for 28 nm embedded STT-MRAM and beyond,"Applied Materials Inc, Santa Clara, CA, US"
+A 0.004mm2 250μW ΔΣ TDC with time-difference accumulator and a 0.012mm2 2.5mW bang-bang digital PLL using PRNG for low-power SoC applications,"Samsung Electronics,Yongin,Korea"
+"Varistor-type bidirectional switch (JMAX>107A/cm2, selectivity∼104) for 3D bipolar resistive memory arrays","Dept. Nanobio Mat. and Elec.,Gwangju Institute of Science and Technology,Korea"
+Control of electro-chemical etching for uniform 0.1 /spl mu/m gate formation of HEMT,"Semiconductor Technology Laboratory, Oki Electric Industry Company Limited, Hachioji, Tokyo, Japan"
+Temperature calibration of CMOS magnetic vector probe for contactless angle measurement system,"Physical Electronics Laboratory, Zurich, Switzerland"
+FinFET-a quasi-planar double-gate MOSFET,"Dept. of Electr. Eng. & Comput. Sci.,California Univ.,Berkeley,CA,USA"
+Designing High Performance Microprocessors,"Digital Equipment Corporation Hudson,Massachusetts USA"
+A 247 and 272 GHz Two-Stage Regenerative Amplifiers in 65 nm CMOS with 18 and 15 dB Gain Based on Double-Gmax Gain Boosting Technique,"Department of Electrical Engineering,KAIST,South Korea; IMEC,Belgium"
+Measurement of Nano-Displacement Based on In-Plane Suspended-Gate MOSFET Detection Compatible with a Front-End CMOS Process,"CEA-LETI,Grenoble,France"
+Megapixel CMOS image sensor fabricated in three-dimensional integrated circuit technology,"Lincoln Lab.,MIT,Lexington,MA,USA"
+A 5500FPS 85GOPS/W 3D Stacked BSI Vision Chip Based on Parallel in-Focal-Plane Acquisition and Processing,"LIST,CEA,Saclay,France"
+Approaching fermi level unpinning in Oxide-In0.2Ga0.8As,"Intel Corporation, Santa Clara, CA, USA"
+Highly manufacturable 90 nm DRAM technology,"Technology Development, Semiconductor Research and Development Division, Samsung Electronics Company Limited, Yongin si, Gyeonggi, South Korea"
+A 0.25 mm x86 microprocessor with a 100 MHz socket 7 interface,"Adv. Micro Devices,Milpitas,CA,USA"
+Cost-effective high-performance high-voltage SiGe:C HBTs with 100 GHz f/sub T/ and BV/sub CEO/ /spl times/ f/sub T/ products exceeding 220 VGHz,"IHP, Frankfurt, Germany"
+An Approach to Embedding Traditional Non-Volatile Memories into a Deep Sub-Micron CMOS,"Taiwan Semiconductor Manufacturing Company,Ltd,Integrated Interconnect & Packaging,R&D,Hsinchu,Taiwan,R.O.C."
+6.2 A 460mW 112Gb/s DSP-Based Transceiver with 38dB Loss Compensation for Next-Generation Data Centers in 7nm FinFET Technology,"MediaTek,Irvine,CA"
+Can InAlN/GaN be an alternative to high power / high temperature AlGaN/GaN devices?,"I.E.M.N, Villeneuve d'Ascq, France"
+A Novel Cross-Spacer Phase Change Memory with Ultra-Small Lithography Independent Contact Area,"ITRI, Material and Chemical Research Laboratories, Hsinchu, Taiwan, R.O.C"
+A 100dB SNR 2.5MS/s output data rate /spl Delta//spl Sigma/ ADC,"Analog Devices,Newbury,UK"
+Industrially Applicable Read Disturb Model and Performance on Mega-Bit 28nm Embedded RRAM,"Quality and Reliability,Taiwan Semiconductor Manufacturing Company,,Park Ave.,Hsinchu Science Park Hsinchu,Taiwan,,R.O.C"
+Competitive and cost effective high-k based 28nm CMOS technology for low power applications,"IBM Semiconductor Research and Development Center (SRDC), STMicroelectronics, Inc., Hopewell Junction, NY, USA"
+Electrical characteristics and reliability of sub-3 nm gate oxides grown on nitrogen implanted silicon substrates,"SRDC, IBM Corp., Hopewell Junction, NY, USA"
+Scaling of Ω-gate SOI nanowire N- and P-FET down to 10nm gate length: Size- and orientation-dependent strain effects,"STMicroelectronics,,rue J. Monnet,Crolles,France"
+Understanding and prediction of EWF modulation induced by various dopants in the gate stack for a gate-first integration scheme,"TSMC,Belgium"
+A VDSL2 CPE AFE in 0.15µm CMOS with integrated line driver,"Marvell,Santa Clara,CA,USA"
+A 230–260GHz wideband amplifier in 65nm CMOS based on dual-peak Gmax-core,"Department of Electrical Engineering,CBNU,South Korea"
+Scaling of 32nm low power SRAM with high-K metal gate,"Samsung Electronics, Hopewell Junction, USA"
+Highly-scalable threshold switching select device based on chaclogenide glasses for 3D nanoscaled memory arrays,"Semiconductor Device Laboratory, Nano Fabrication Group, Samsung Advanced Institute of Technology, Gyeonggi-Do, Korea"
+Selectively formed high mobility strained Ge PMOSFETs for high performance CMOS,"Systems and Technology Group, Hopewell Junction, NY, USA"
+A digitally calibrated 5.15-5.825GHz transceiver for 802.11a wireless LANs in 0.18/spl mu/m CMOS,"Athena Semiconductors,Fremont,CA,USA"
+A 43mW Bluetooth transceiver with -91dBm sensitivity,"Skyworks Solutions,Ottawa,Ont.,Canada"
+OC-192 transmitter in standard 0.18 /spl mu/m CMOS,"Broadcom Corp.,Irvine,CA,USA"
+High-performance InSb based quantum well field effect transistors for low-power dissipation applications,"QinetiQ, Malvern Technology Centre, Malvern, UK"
+27.2 A 6mW 5K-Word real-time speech recognizer using WFST models,"Massachusetts Institute of Technology,Cambridge,MA"
+A 20mW 85dB/spl Omega/ 1.25Gb/s CMOS transimpedance amplifier with photodiode capacitance cancellation,"SoC Technol. Center,Ind. Technol. Res. Inst.,Hsinchu,Taiwan"
+A Silicon Photonics Technology for 400 Gbit/s Applications,"STMicroelectronics, Agrate, Italy"
+"Design and process integration for high-density, high-speed, and low-power 6F/sup 2/ cross point MRAM cell","Corporate Research & Development Center, Toshiba Corporation, Kanagawa, Japan"
+GaN-based Periodic High-Q RF Acoustic Resonator with Integrated HEMT,"the US Naval Research Laboratory, National Research Council Fellow residing, Washington DC, USA"
+Advanced interconnect schemes towards 0.1 /spl mu/m,"LETI-CEA Technologies Avancees, Grenoble, France"
+"Technology Breakthrough of Low Temperature, Low Defect, and Low Cost SiGe Selective Epitaxial Growth (L3 SiGe SEG) Process for 45nm Node and Beyond","Hitachi Kokusai Electric Inc.,Yasuuchi,Yatsuo-machi,Toyama,Japan."
+Applications and design styles for 3DIC,"Synopsys Inc., Santa Clara, CA, USA"
+Application and Benefits of Target Programming Algorithms for Ferroelectric HfO2 Transistors,"Ferroelectric Memory GmbH,Dresden,Germany"
+22.6 A 22V compliant 56µW active charge balancer enabling 100% charge compensation even in monophasic and 36% amplitude correction in biphasic neural stimulators,"Hahn-Schickard,Villingen-Schwenningen,Germany"
+A Stacked Embedded DRAM Array for LPDDR4/4X using Hybrid Bonding 3D Integration with 34GB/s/1Gb 0.88pJ/b Logic-to-Memory Interface,"Wuhan Xinxin Semiconductor Manufacturing Co.,Ltd.,Wuhan,China"
+A new vertically stacked poly-Si MOSFET for 533 MHz high speed 64Mbit SRAM,"Hitachi Cambridge Laboratory, Hitachi Europe Ltd., Cambridge, UK"
+High electron and hole mobility enhancements in thin-body strained Si/strained SiGe/strained Si heterostructures on insulator,"Department of Materials Science and Engineering, MIT, Cambridge, MA, USA"
+Low operation voltage high integrated field emitter arrays by transfer metal mold technique using ultra precision machining and super microelectroplating technology,"Corporate Research & Development Center, Toshiba Corporation, Kawasaki, Japan"
+High performance CMOS fabricated on hybrid substrate with different crystal orientations,"Microelectronic Division, Hopewell Junction, NY, USA"
+Advanced power devices for many-core processor power supplies,"ACOO Enterprises LLC, USA"
+"A Low-Cost, High-Performance, High-Voltage Complementary BiCMOS Process","Im Technologiepark, IHP, Frankfurt, Germany"
+High performance poly-Si TFTs on a glass by a stable scanning CW laser lateral crystallization,"Fujitsu Laboratories Limited, Atsugi, Japan"
+Role of correlation in systematic variation modeling,"Compact Device Modeling Group, Advanced Design, Intel Corporation, Hillsboro, US"
+Role of temperature in process-induced charging damage in sub-micron CMOS transistors,"Sematech, Austin, TX, USA"
+A SiGe transmitter chipset for CATV video-on-demand systems,"Microtune,Plano,TX,USA"
+Surface Wave and Lamb Wave Acoustic Devices on Heterogenous Substrate for 5G Front-Ends,"Harbin Institute of Technology,School of Science,Shenzhen,China"
+Managing leakage in charge-based analog circuits with low-V/sub TH/ transistors by analog T-switch (AT-Switch) and super cut-off CMOS,"Center for Collaborative Res.,Tokyo Univ.,Japan"
+On the dynamic resistance and reliability of phase change memory,"IBM Hopewell Junction,USA"
+A novel W/WNx/dual-gate CMOS technology for future high-speed DRAM having enhanced retention time and reliability,"Technology & Development Office, Elpida Memory, Inc., Sagamihara, Kanagawa, Japan"
+300mm Heterogeneous 3D Integration of Record Performance Layer Transfer Germanium PMOS with Silicon NMOS for Low Power High Performance Logic Applications,"Components Research, Intel Corporation, Hillsboro, OR, USA"
+0.18 um modular triple self-aligned embedded split-gate flash memory,"Div. of Microelectron.,IBM,Hopewell Junction,NY,USA"
+ESD Protection for Mixed-Voltage I/O in LowVoltage Thin-Oxide CMOS,"Nat. Chiao-Tung Univ.,Hsin-Chu"
+A 256MB synchronous-burst DDR SRAM with hierarchical bit-line architecture for mobile applications,"Samsung,Hwasung,South Korea"
+A low power and high speed data transfer scheme with asynchronous compressed pulse width modulation for AS-memory,"ULSI Lab.,Mitsubishi Electr. Corp.,Itami,Japan"
+A CMOS Image Sensor Integrating Column-Parallel Cyclic ADCs with On-Chip Digital Error Correction Circuits,"Sanei Hytechs,Hamamatsu,Japan"
+An orthogonal 6F/sup 2/ trench-sidewall vertical device cell for 4 Gb/16 Gb DRAM,"Infineon Technologies, Dresden, Germany"
+A middle-1X nm NAND flash memory cell (M1X-NAND) with highly manufacturable integration technologies,"Research and Development Division, Flash Device development & Advanced Process Team, Ichon, Gyeonggi, South Korea"
+A 400MHz random-cycle dual-port interleaved DRAM with striped-trench capacitor,"Matsushita,Nagaokakyo,Japan"
+A 1mW Dual-Chopper Amplifier for a 50-/spl mu/g/spl radic/Hz Monolithic CMOS-MEMS Capacitive Accelerometer,"Dept. of Electr. & Comput. Eng.,Florida Univ.,Gainesville,FL"
+A large-area curved pyroelectric fingerprint sensor,"TNO Holst Centre, The Netherlands"
+A 4.5GHz LC-VCO with Self-Regulating Technique,"Renesas Technology,Takasaki,Japan"
+Advanced MMIC for Passive Millimeter and Submillimeter Wave Imaging,"Northrop Grumman,Redondo Beach,CA"
+Random Telegraph Signal Statistical Analysis using a Very Large-scale Array TEG with 1M MOSFETs,"Asahi Kasei Microsystems Co.,Ltd.,Aza-Aoba,Aramaki,Aoba-ku,Sendai,,Japan"
+High frequency InAs-channel HEMTs for low power ICs,"HRL Laboratories LLC, CA, USA"
+Ultra thinning 300-mm wafer down to 7-µm for 3D wafer Integration on 45-nm node CMOS using strained silicon and Cu/Low-k interconnects,"Fujitsu Laboratories Limited, Atsugi, Kanagawa, Japan"
+A 7nm FinFET technology featuring EUV patterning and dual strained high mobility channels,"GLOBALFOUNDRIES, Albany Nanotechnology Center, Albany, NY"
+PBTI/NBTI monitoring ring oscillator circuits with on-chip Vt characterization and high frequency AC stress capability,"IBM SRDC,Hopewell Junction,NY,USA"
+SOI circuit technology for batteryless mobile system with green energy sources,"Commun. Device R&D Dept.,Seiko Epson Corp.,Nagano,Japan"
+A 10Gb/s eye-opening monitor in 0.13 /spl mu/m CMOS,"California Inst. of Technol.,Pasadena,CA,USA"
+"High performance low temperature activated devices and optimization guidelines for 3D VLSI integration of FD, TriGate, FinFET on insulator","STMicroelectronics,France"
+Advances in 3D CMOS sequential integration,"CEA, MINATEC, Grenoble, France"
+Hybrid 1T e-DRAM and e-NVM Realized in One 10 nm node Ferro FinFET device with Charge Trapping and Domain Switching Effects,"Key Laboratory of Microelectronics Devices and Integrated Technology, Chinese Academy of Sciences, Beijing, China"
+Advanced power electronic devices based on Gallium Nitride (GaN),"Cambridge Electronics, Inc. (CEI), Cambridge, MA, USA"
+0.1 /spl mu/m level contact hole pattern formation with KrF lithography by resolution enhancement lithography assisted by chemical shrink (RELACS),"Ryoden Semiconductor System Engineering Corporation, Itami, Hyogo, Japan"
+Weak inversion MOS varactors for 0.5 V analog integrated filters,"Columbia Univ.,New York,NY,USA"
+Analysis and control of hysteresis in PD/SOI CMOS,"IBM Research Division, Yorktown Heights, NY, USA"
+"19.6 A 0.2V trifilar-coil DCO with DC-DC converter in 16nm FinFET CMOS with 188dB FOM, 1.3kHz resolution, and frequency pushing of 38MHz/V for energy harvesting applications","1TSMC,Hsinchu,Taiwan"
+A 0.18 /spl mu/m CMOS front-end processor for a blu-ray disc recorder with an adaptive PRML,"Samsung Electron.,Suwon,South Korea"
+"A 90-nm CMOS device technology with high-speed, general-purpose, and low-leakage transistors for system on chip applications","Taiwan Semiconductor Manufacturing Company, Science Based Industrial Park, Hsinchu, Taiwan"
+18.3 A 120mA Non-Isolated Capacitor-Drop AC/DC Power Supply,"Texas Instruments,Tucson,AZ"
+Novel SiC power MOSFET with integrated unipolar internal inverse MOS-channel diode,"Advanced Devices Development Center, Panasonic Corporation, Moriguchi, Osaka, Japan"
+Wireless implantable microsystems: coming breakthroughs in health care,"Dept. of Electr. Eng. & Comput. Sci.,Michigan Univ.,Ann Arbor,MI,USA"
+High temperature operation of AlInAs/InGaAs/AlInAs 3D-SMODFETs with record two-dimensional electron gas densities,"US Army Research Laboratory, Fort Monmouth, NJ, USA"
+A highly manufacturable low-k ALD-SiBN process for 60nm NAND flash devices and beyond,"Process Engineering Section, Thermal Processing Systems BU, Tokyo Electron Limited, Nirasaki, Yamanashi, Japan"
+A 28nm HKMG super low power embedded NVM technology based on ferroelectric FETs,"NaMLab gGmbH, Dresden, Germany"
+Gate-all-around Twin Silicon nanowire SONOS Memory,"PD Team,San,Nongseo-Dong,Kiheung-Ku,Yongin-City,Kyoungi-Do,,KOREA"
+Quantized conductive filament formed by limited Cu source in sub-5nm era,"School of Materials Science and Engineering, Department of Nanobio Materials and Electronics, Gwangju Institute of Science and Technology, Gwangju, South Korea"
+Additive manufacturing for electronics “Beyond Moore”,"PARC, A Xerox company, Palo Alto, USA"
+SRAM current-sense amplifier with fully-compensated bit line multiplexer,"Tech. Univ. of Munich,Germany"
+Performance comparison of sub 1 nm sputtered TiN/HfO/sub 2/ nMOS and pMOSFETs,"Texas Instruments, USA"
+2RW dual-port SRAM design challenges in advanced technology nodes,"Renesas System Design Corporation, Tokyo, Japan"
+Experimental characterization of stiction due to charging in RF MEMS,"E.E. Department of K. U. Leuven, IMEC vzw, Leuven, Belgium"
+A 4.75GHz fractional frequency divider with digital spur calibration in 45nm CMOS,"Intel,Hillsboro,OR,USA"
+A wire-speed powerTM processor: 2.3GHz 45nm SOI with 16 cores and 64 threads,"IBM Research,Bedford,NH,USA"
+"GaN Power Commercialization with Highest Quality-Highest Reliability 650V HEMTs-Requirements, Successes and Challenges","Transphorm Inc., Goleta, CA, USA"
+A 0.5-to-480MHz Self-Referenced CMOS Clock Generator with 90ppm Total Frequency Error and Spread-Spectrum Capability,"Mobius Microsystems,Detroit,MI"
+0.228 /spl mu/m/sup 2/ trench cell technologies with bottle-shaped capacitor for 1 Gbit DRAMs,"ULSI Research Laboratories, Toshiba Corporation, Kawasaki, Japan"
+A 14nm FinFET transistor-level 3D partitioning design to enable high-performance and low-cost monolithic 3D IC,"Technology Development, GLOBALFOUNDRIES, Malta, Ny, USA"
+Physics-based compact modeling framework for state-of-the-art and emerging STT-MRAM technology,"Device Lab, Samsung Semiconductor Inc., San Jose, CA, USA"
+Anomalous diffusion in the extension region of nanoscale MOSFETs,"FUJITSU LABORATORIES Limited, Atsugi, Kanagawa, Japan"
+Enabling UTBB Strained SOI Platform for Co-Integration of Logic and RF: Implant-Induced Strain Relaxation and Comb-Like Device Architecture,"CEA,LETI,Minatec Campus,Grenoble,France"
+A 5-mW 6-Gb/s Quarter-Rate Sampling Receiver with a 2-Tap DFE Using Soft Decisions,"California Univ.,Los Angeles,CA"
+A new direct low-k/Cu dual damascene (DD) contact lines for low-loss (LL) CMOS device platforms,"NEC Electronics Corporation,Shimokuzawa,Sagamihara,Kanagawa,JAPAN"
+"First fully functionalized monolithic 3D+ IoT chip with 0.5 V light-electricity power management, 6.8 GHz wireless-communication VCO, and 4-layer vertical ReRAM","National Nano Device Laboratories, Hsinchu, Taiwan"
+A −31dBc integrated-phase-noise 29GHz fractional-N frequency synthesizer supporting multiple frequency bands for backward-compatible 5G using a frequency doubler and injection-locked frequency multipliers,"FCI,Seongnam,Korea"
+A Spur Suppression Technique for Phase-Locked Frequency Synthesizers,"National Taiwan Univ.,Taipei"
+Experimental and comparative investigation of low and high field transport in substrate- and process-induced strained nanoscaled MOSFETs,"CEA-LETI/DRT,Grenoble,France"
+III-V HEMTs for Cryogenic Low Noise Amplifiers,"Low Noise Factory AB,Gothenburg,Sweden"
+A 14b 40MS/s Redundant SAR ADC with 480MHz Clock in 0.13pm CMOS,"Infineon Technologies,Munich,Germany"
+A system-on-chip for bi-directional point-to-multipoint wireless digital audio applications,"Catena,Kista,Sweden"
+High performance amorphous oxide thin film transistors with self-aligned top-gate structure,"Semiconductor Laboratory, Samsung Advanced Institute of Technology, Yongin si, Gyeonggi, South Korea"
+Design and fabrication of a high dynamic range image sensor in TFA technology,"Inst. fur Halbleiterelektronik,Siegen Univ.,Germany"
+Circuit yield of organic integrated electronics,"STMicroelectronics,Milan,Italy"
+A 0.25 mW sigma-delta modulator for voice-band applications,"Integrated Syst. Labs.,Texas Instrum. Inc.,Dallas,TX,USA"
+PRD-based global-mean-time signaling for high-speed chip-to-chip communications,"Fujitsu Labs. Ltd.,Atsugi,Japan"
+Hybrid silicon/molecular memories: co-engineering for novel functionality,"ZettaCore, Inc., CO, USA"
+A Formation of Si Native Oxide Membrane Using High-Selectivity Etching and Applications for Nano-Pipe Array and Micro-Diaphragm on Si Substrate,"Spansion, Inc., Sunnyvale, CA, USA"
+A harmonic rejection mixer robust to RF device mismatches,"Silicon Laboratories,Austin,TX"
+Monolithically integrated 600-V E/D-mode SiNx/AlGaN/GaN MIS-HEMTs and their applications in low-standby-power start-up circuit for switched-mode power supplies,"Department of Microwave Devices and IC's, Institute of Microelectronics, Beijing, China"
+"A 0.016mm2, 2.4GHz RF signal quality measurement macro for RF test and diagnosis","System Devices Research Laboratories,NEC Corporation,Sagamihara,Kanagawa,,Japan"
+The 300mm Technology Current Status And Future Prospect,"Semiconductor Leading Edge Technologies,Inc. Yoshida+ho,Totsuka-ku,Yokohama-shi,Japan"
+3.6 A 6-to-600MS/s Fully Dynamic Ringamp Pipelined ADC with Asynchronous Event-Driven Clocking in 16nm,"imec,Leuven,Belgium"
+10.5 A Fully Integrated 27dBm Dual-Band All-Digital Polar Transmitter Supporting 160MHz for WiFi 6 Applications,"Intel,Haifa,Israel"
+Epitaxial strained germanium p-MOSFETs with HfO/sub 2/ gate dielectric and TaN gate electrode,"Intel Corporation, Hillsboro, OR, USA"
+"Fully integrated 1.7GHz, 188dBc/Hz FoM, 0.8V, 320/spl mu/W LC-tank VCO and frequency divider","Center for Phys. Electron.,Denmark Tech. Univ.,Lyngby,Denmark"
+An 8640 MIPS SoC with Independent Power-Off Control of 8 CPUs and 8 RAMs by An Automatic Parallelizing Compiler,"Hitachi,Tokyo,Japan"
+A Wireless Transceiver with Integrated Data Converters for 802.11a/b/g Access Points,"Analog Devices,Raleigh,NC"
+A low-power integrated tuner for cable-telephony applications,"Silicon Wave Inc.,San Diego,CA,USA"
+"A 3MHz-BW 3.6GHz digital fractional-N PLL with sub-gate-delay TDC, phase-interpolation divider, and digital mismatch cancellation","Politecnico di Milano,Italy"
+A thin amorphous silicon buffer process for suppression of W polymetal gate depletion in PMOS,"T Project Group,Fujitsu Labs. Ltd.,Japan"
+A single-chip CMOS transceiver for DCS-1800 wireless communications,"Katholieke Univ.,Leuven,Belgium"
+A novel solution for porous low-k dual damascene post etch stripping/clean with supercritical CO/sub 2/ technology for 65nm and beyond applications,"Taiwan Semiconductor Manufacturing Company Limited, Hsinchu, Taiwan"
+Damage-free CMP towards 32nm-node porous low-k (k = 1.6)/Cu integration,"Semicond. Leading Edge Technol. Inc.,Ibaraki,Japan"
+A new cell structure for sub-quarter micron high density flash memory,"VLSI. Research Laboratory, Sharp Corporation, Nara, Japan"
+Coupled quantum dots on SOI as highly integrated Si qubits,"Department of Electrical Engineering, Tokyo Institute of Technology, Tokyo, Japan"
+10.6 A 4G/5G Cellular Transmitter in 12nm FinFET with Harmonic Rejection,"MediaTek,Kent,United Kingdom"
+Channel Stress Modulation and Pattern Loading Effect Minimization of Milli-Second Super Anneal for Sub-65nm High Performance SiGe CMOS,"Res. & Dev.,Taiwan Semicond. Manuf. Co. Ltd.,Hsinchu"
+An Internally-matched GaN HEMT Amplifier with 550-watt Peak Power at 3.5 GHz,"Cree Research, Inc., Goleta, CA, USA"
+A 14-bit 8.9GS/s RF DAC in 40nm CMOS achieving >71dBc LTE ACPR at 2.9GHz,"Texas Instruments Incorporated,Dallas,USA"
+Intrinsic fluctuations in Vertical NAND flash memories,"Process Development Team,Semiconductor R&D Center,Samsung Electronics Co. Ltd.,,Banwol-Dong,Hwasung-City,Gyunggi-Do,Korea"
+Highly scalable flash memory with novel deep trench isolation embedded into highperformance cmos for the 90nm node & beyond,"Infineon Technologies NA, NY, USA"
+Effect of mechanical stress on reliability of gate-oxide film in MOS transistors,"Mechanical Engineering Research Laboratory, Hitachi and Limited, Tsuchiura, Ibaraki, Japan"
+An 8Mb demonstrator for high-density 1.8V Phase-Change Memories,"MPG & Central R&D,STMicroelectronics,Agrate Brianza,Italy"
+A 375 MHz 1 /spl mu/m CMOS 8-bit multiplier,"Integrated Syst. Lab.,Swiss Federal Inst. of Technol.,Zurich,Switzerland"
+20 Gb/s self-timed vector processing with Josephson single-flux quantum technology,"IBM Austin Research Laboratory,Austin,TX,USA"
+Fully-parallel 25 MHz 2.5 Mb CAM,"Nortel Semiconductors,Ottawa,Ont.,Canada"
+On the microscopic origin of the frequency dependence of hole trapping in pMOSFETs,"Imec, Leuven, Belgium"
+A 50-nm 1.2-V GexTe1−x/Sb2Te3 superlattice topological-switching random-access memory (TRAM),"Low-power Electronics Association & Project,Onogawa,Tsukuba,Ibaraki,JAPAN"
+A Current Driver IC using a S/H for QVGA FullColor Active-Matrix Organic LED Mobile Displays,"Samsung Electronics,Yong-In City,Korea"
+Poly pitch and standard cell co-optimization below 28nm,"ARM INC, Austin, TX, USA"
+The effect of interconnect scaling and low-k dielectric on the thermal characteristics of the IC metal,"Semiconductor Process and Device Center, Texas Instruments, Inc., Dallas, TX, USA"
+Comprehensive extensibility of 20nm low power/high performance technology platform featuring scalable high-k/metal gate planar transistors with reduced design corner,"Samsung Electronics Company Limited, Yongin, Gyeonggi, South Korea"
+Precursor ISI Reduction in High-Speed I/O,"Rambus,Inc,Los Altos,CA; MIT,Cambridge,MA"
+First Transistor Demonstration of Thermal Atomic Layer Etching: InGaAs FinFETs with sub-5 nm Fin-width Featuring in situ ALE-ALD,"MIT, Microsystems Technology Laboratories, Cambridge, MA, USA"
+Correlation of low-frequency noise and emitter-base reverse-bias stress in epitaxial Si- and SiGe-base bipolar transistors,"IBM Microelectronics, Hopewell Junction, NY, USA"
+Re-Examination of Vth Window and Reliability in HfO2 FeFET Based on the Direct Extraction of Spontaneous Polarization and Trap Charge during Memory Operation,"Kioxia Corporation,Institute of Memory Technology Research & Development,Yokkaichi,Japan"
+Physical mechanisms of endurance degradation in TMO-RRAM,"A*STAR, Institute of Microelectronics, Singapore, Singapore"
+Demonstration of recessed SiGe S/D and inserted metal gate on HfO/sub 2/ for high performance pFETs.,"TI assignee at IMEC, Heverlee, Belgium"
+A 3.6GB/s 1.3mW 400mV 0.051mm2 near-threshold voltage resilient router in 22nm tri-gate CMOS,"SoC Design Lab,Intel Labs,Intel Corporation,Hillsboro,OR,USA"
+Ultrathin (<10nm) Nb2O5/NbO2 hybrid memory with both memory and selector characteristics for high density 3D vertically stackable RRAM applications,"School of Materials Science and Engineering,Gwangju Institute of Science and Technology,Korea"
+A novel semimetallic quantum well FET,"Naval Research Laboratory, Inc., Washington D.C., DC, USA"
+"A 4.6μm, 512×512, Ultra-Low Power Stacked Digital Pixel Sensor with Triple Quantization and 127dB Dynamic Range","Brillnics Japan Inc.,Tokyo,Japan"
+A 1 1/4 inch 8.3M pixel digital output CMOS APS for UDTV application,"Micron Technology,CA,USA"
+An Approach to Embedding Traditional Non-Volatile Memories into a Deep Sub-Micron CMOS,"ATQRD,TSMC,Hsinchu,Taiwan"
+Combined linear-logarithmic CMOS image sensor,"Edinburgh Univ.,UK"
+Programmable and automatically-adjustable sense-amplifier activation scheme and multi-reset address-driven decoding scheme for high-speed reusable SRAM core,"Device Dev. Center,Hitachi Ltd.,Tokyo,Japan"
+14nm FDSOI technology for high speed and energy efficient applications,"IBM,rue Jean Monnet,Crolles,France"
+The impact of substrate surface potential on the performance of RF power LDMOSFETs on high-resistivity SOI,"AmberWave Systems Corporation, Salem, NH, USA"
+Novel fabrication process and structure of a low-voltage-operation micromirror array for optical MEMS switches,"NTT Microsystem Integration Laboratories, Japan"
+Flexible and robust capping-metal gate integration technology enabling multiple-VT CMOS in MuGFETs,"IMEC,Belgium"
+A record GmSAT/SSSAT and PBTI reliability in Si-passivated Ge nFinFETs by improved gate stack surface preparation,"imec,Kapeldreef,Leuven,,Belgium"
+High thermal tolerance of 25-nm c-axis aligned crystalline In-Ga-Zn oxide FET,"Semiconductor Energy Laboratory Co., Ltd., Atsugi, Kanagawa, Japan"
+A quad band WCDMA transceiver with fractional local divider,"Hitachi Central Research Laboratory,Japan"
+"A 2.0 V, 0.35 /spl mu/m partially depleted SOI-CMOS technology","Digital Equipment Corporation, Hudson, MA, USA"
+A 200 MSample/s trellis-coded PRML read/write channel with digital servo,"SGS-Thomson Microelectronics,San Jose,CA,USA"
+Technology Innovations In Mobile Computers,"IBM Japan,Shtmots-uruma,,Kanagawa-!ten,Japan"
+A 4.25 GHz BiCMOS clock recovery circuit with an AV-DSPD architecture for NRZ data stream,"NEC Corp.,Kanagawa,Japan"
+A novel poly-silicon-capped poly-silicon-germanium thin-film transistor,"Xerox Palo Alto Research Center, Palo Alto, CA"
+"A 14.7Mb/mm2 28nm FDSOI STT-MRAM with Current Starved Read Path, 52Ω/Sigma Offset Voltage Sense Amplifier and Fully Trimmable CTAT Reference","ARM,San Jose,CA,USA"
+A Self-Resonant MEMS-based Electrostatic Field Sensor with 4V/m/Hz Sensitivity,"Medtronic,Fridley,MN"
+Record high mobility (428cm2/V-s) of CVD-grown Ge/strained Ge0.91Sn0.09/Ge quantum well p-MOSFETs,Graduate Institute of Electronics Engineering
+1st quantitative failure-rate calculation for the actual large-scale SRAM using ultra-thin gate-dielectric with measured probability of the gate-current fluctuation and simulated circuit failure-rate,"Fujitsu Ltd.,,Fuchigami,Akiruno,Tokyo,Japan"
+A 10MHz 80μW 67 ppm/°C CMOS reference clock oscillator with a temperature compensated feedback loop in 0.18μm CMOS,"Department of EECS,KAIST,Guseong-dong,Yuseong-gu,Daejon,Republic of Korea"
+Strain engineered extremely thin SOI (ETSOI) for high-performance CMOS,"GLOBALFOUNDRIES,Albany,NY,USA"
+A fully integrated zero-IF transceiver for GSM-GPRS quad band application,"Texas Instruments,Villeneuve Loubet,France"
+Comprehensive study on AC characteristics in SOI MOSFETs for analog applications,"Dept. of Electr. Eng.,California State Univ.,Los Angeles,CA,USA"
+Micro-Engineered Devices for Motion Energy Harvesting,"Department of Electrical & Electronic Engineering, Imperial College London, London, UK"
+Novel self-assembled ultra-low-k porous silica films with high mechanical strength for 45 nm BEOL technology,"MIRAI, Association of Super-Advanced Electronics Technologies (ASET), Japan"
+An Analog Frontend Chip for a MEMS-Based Parallel Scanning-Probe Data-Storage System,"IBM Systems & Technology Group,Essex Junction,Vermont"

data/semiconductor_org_types/train.csv

@@ -0,0 +1,51 @@
+Paper title,Organization name,Label
+3Gb/s AC-coupled chip-to-chip communication using a low-swing pulse receiver,"North Carolina State Univ.,Raleigh,NC,USA",university
+Sub-Micron CMOS / MOS-Bipolar Hybrid TFTs for System Displays,"Advanced LCD Technology Development Center Company Limited, Yokohama, Kanagawa, Japan",company
+24.4 A 680nA fully integrated implantable ECG-acquisition IC with analog feature extraction,"imec,Heverlee,Belgium",research institute
+A write-back cache memory using bit-line steal technique,"Corp. Semicond. Dev. Div.,Matsushita Electr. Ind. Co. Ltd.,Kyoto,Japan",company
+High performance 0.25 /spl mu/m gate-length doped-channel AlGaN/GaN heterostructure field effect transistors grown on p-type SiC substrates,"APA Optics, Inc., Blaine, MN, USA",company
+3-terminal nanoelectromechanical switching device in insulating liquid media for low voltage operation and reliability improvement,"National NanoFab Center, Daejeon, South Korea",research institute
+Full metal gate with borderless contact for 14 nm and beyond,"Toshiba at Albany NanoTech,NY,USA",company
+A novel self-aligned shallow trench isolation cell for 90 nm 4 Gbit NAND flash EEPROMs,"SoC R&D Center,Semiconductor Company,Toshiba Corp.,Isogo-ku,Yokohama,Japan",company
+A 0.13/spl mu/m CMOS EDGE/UMTS/WLAN Tri-Mode /spl Delta//spl Sigma/ ADC with -92dB THD,"ETH,Zurich,Switzerland; Advanced Circuit Pursuit,Zollikon,Switzerland",university
+On the gate oxide scaling of high performance CMOS transistors,"Semiconductor R&D Center, Samsung Electronics Co., Ltd, Yongin-City, Gyeonggi-Do, Korea (ROK)",company
+A 0.13/spl mu/m CMOS EDGE/UMTS/WLAN Tri-Mode /spl Delta//spl Sigma/ ADC with -92dB THD,"Advanced Circuit Pursuit,Zollikon,Switzerland; ETH,Zurich,Switzerland",company
+A 3Gb/s 8b single-ended transceiver for 4-drop DRAM interface with digital calibration of equalization skew and offset coefficients,"Pohang Univ. of Sci. & Technol.,South Korea",university
+25.5 A Self-Calibrated 1.2-to-3.8GHz 0.0052mm2 Synthesized Fractional-N MDLL Using a 2b Time-Period Comparator in 22nm FinFET CMOS,"Intel,Hillsboro,OR",company
+Accurate performance evaluation for the horizontal nanosheet standard-cell design space beyond 7nm technology,"GLOBALFOUNDRIES Inc., Albany, NY, USA",company
+"Front-end-of-line (FEOL) optimization for high-performance, high-reliable strained-Si MOSFETs; from virtual substrate to gate oxidation","Memory Division, Samsung Electronics Co, Yongin-City, Gyeonggi-Do, Korea",company
+A 14 b 100 Msample/s CMOS DAC designed for spectral performance,"Illinois Univ.,Urbana,IL,USA",university
+Design of the Power6 Microprocessor,"IBM Systems Group,Austin,TX",company
+Collective-effect state variables for post-CMOS logic applications,"Strategic Technology Group,Advanced Micro Devices,Sunnyvale,CA,USA",company
+Single-chip IF transceiver IC with wide dynamic range variable gain amplifiers for wideband CDMA applications,"Syst. LSI Dev. Center,Mitsubishi Electr. Corp.,Hyogo,Japan",company
+Formation of Si-on-Insulator Structure by Lateral Solid Phase Epitaxial Growth with Local P-Doping,"Central Research Laboratory,Hitachi Ltd. Kokubunji. Tokyo,Japan",company
+1D thickness scaling study of phase change material (Ge2Sb2Te5) using a pseudo 3-terminal device,"Samsung Electronics Company Limited, Yongin si, Gyeonggi, South Korea",company
+A 500MHz multi-banked compilable DRAM macro with direct write and programmable pipelining,"IBM Microelectron.,Burlington,VT,USA",company
+Dislocation engineering for a silicon-based light emitter at 1.5 /spl mu/,"MPI für Mikrostrukturphysik, Halle, Germany",research institute
+An enhanced 130 nm generation logic technology featuring 60 nm transistors optimized for high performance and low power at 0.7 - 1.4 V,"QRE, Hillsboro, OR, USA",company
+Physical understanding of Vth and Idsat variations in (110) CMOSFETs,"Center for Semiconductor Research & Development,Toshiba Corporation,Japan",company
+Destructive-read random access memory system buffered with destructive-read memory cache for SoC applications,"IBM Microelectron.,Hopewell Junction,NY,USA",company
+Benchmarking of monolithic 3D integrated MX2 FETs with Si FinFETs,"KUL, Leuven, Belgium",university
+A 48-mW 18-Gb/s fully integrated CMOS optical receiver with photodetector and adaptive equalizer,"Applied Science and Technology Research Institute,Hong Kong",research institute
+Role of non-radiative recombination in the degradation of InGaN-based laser diodes,"Matsushita Electric Industrial Limited, Takatsuki, Osaka, Japan",company
+Highly area efficient and cost effective double stacked S/sup 3/ (stacked single-crystal Si) peripheral CMOS SSTFT and SRAM cell technology for 512M bit density SRAM,"R & D Center, Samsung Electronics Kiheung-Eup, Yongin-City, Kyungki-do, Korea",company
+"Strained SOI technology for high-performance, low-power CMOS applications","MIRAI-ASET,Kawasaki,Japan",university
+Damascene integration of copper and ultra-low-k xerogel for high performance interconnects,"Texas Instruments Inc, Dallas, TX, US",company
+A crossing charge recycle refresh scheme with a separated driver sense-amplifier for Gb DRAMs,"ULSI Device Dev. Labs.,NEC Corp.,Kanagawa,Japan",company
+Large-signal performance of high-BV/sub CEO/ graded epi-base SiGe HBTs at wireless frequencies,"IBM Microelectronics, Burlington, VT, USA",company
+"A 65 nm CMOS technology with a high-performance and low-leakage transistor, a 0.55 /spl mu/m/sup 2/ 6T-SRAM cell and robust hybrid-ULK/Cu interconnects for mobile multimedia applications","Fujitsu Laboratories Ltd., Atsugi, Kanagawa, Japan",company
+Low-power embedded ReRAM technology for IoT applications,"Incubation Center,Renesas Electronics Corp.,Shimokuzawa,Chuou-ku,Sagamihara,Japan",company
+A DSL customer-premise equipment modem SoC with extended reach/rate for broadband bridging and routing,"Texas Instruments Bangalore and Texas Instruments,Dallas,TX",company
+"An Artificial Iris ASIC with High Voltage Liquid Crystal Driver, 10 nA Light Range Detector and 40 nA Blink Detector for LCD Flicker Removal","Imec,Leuven,Belgium",research institute
+First Demonstration of Low Temperature (≤500°C) CMOS Devices Featuring Functional RO and SRAM Bitcells toward 3D VLSI Integration,"imec from Samsung Electronics,Korea",company
+Competitive and cost effective high-k based 28nm CMOS technology for low power applications,"IBM Semiconductor Research and Development Center (SRDC), Samsung Electronics Company Limited, Hopewell Junction, NY, USA",company
+"Scalable 3D-FPGA using wafer-to-wafer TSV interconnect of 15 Tbps/W, 3.3 Tbps/mm2","Technology Research Department,Association of Super-Advanced Electronics Technologies (ASET),,Higashi-koigakubo,Kokubunji,Tokyo,,Japan",research institute
+"21.8 An all-in-one (Qi, PMA and A4WP) 2.5W fully integrated wireless battery charger IC for wearable applications","MAPS,Yongin,Korea",company
+High performance and low leakage current InGaAs-on-silicon FinFETs with 20 nm gate length,"Samsung Advanced Logic Lab,Austin,TX",company
+Interconnect Scaling Scenario Using A Chip Level Interconnect Model,"Semiconductor Research Center,Matsushita Electric Industrial Co.,Ltd.,Yagumo-nakamachi,Moriguchi,Osaka,Japan",company
+A 12 b 50 M sample/s cascaded folding and interpolating ADC,"Philips Composants et Semiconducteurs,Caen,France",company
+Development of sub 10-µm ultra-thinning technology using device wafers for 3D manufacturing of terabit memory,"Fujitsu Laboratories Ltd.,Japan",company
+A 3.1 to 5 GHz CMOS DSSS UWB transceiver for WPANs,"Sony,Tokyo,Japan",company
+30.1 8b Thin-film microprocessor using a hybrid oxide-organic complementary technology with inkjet-printed P2ROM memory,"Panasonic,Osaka,Japan",company
+Characterizing Electromigration Effects in a 16nm FinFET Process Using a Circuit Based Test Vehicle,"Cisco Systems, Hong Kong, China",company
+"A 180MS/s, 162Mb/s wideband three-channel baseband and MAC processor for 802.11 a/b/g","Engim,Acton,MA,USA",company

data/systematic_review_inclusion/task.json

@@ -0,0 +1 @@
+{"name": "systematic_review_inclusion", "description": "", "data_columns": ["Title", "Abstract", "Authors", "Journal"], "label_columns": {"Label": ["included", "not included"]}}

data/systematic_review_inclusion/train.csv

@@ -0,0 +1,51 @@
+Title,Abstract,Authors,Journal,Label
+Prototyping and transforming facial textures for perception research,Wavelet based methods for prototyping facial textures for artificially transforming the age of facial images were described. Prototype images were used to define the salient features of a particular face classification. Two experiments were conducted to validate the ability of wavelet processing method to capture age information. The first experiment validated the textured prototyping method while the second experiment investigated the effectiveness of the new age transformation technique. The shape and color transformation used to rejuvenate faces hardly affected the apparent age. The average hair color change during rejuvenation was not sufficient to project the hair color in normal range for the younger age group.,"Tiddeman, B.; Burt, M.; Perrett, D.",IEEE Comput Graphics Appl,not included
+School finance reform and voluntary fiscal federalism,"California has transferred the financing of its public schools from localities to the state. In response, many families have supplemented the tax revenue of their local public schools with voluntary contributions. This paper analyzes that phenomenon. We propose a model of partial cooperation among parents in making voluntary contributions to their public schools. Under reasonable conditions, the model predicts that contributions per pupil should decline with school size. We estimate this relationship using data on contributions to California schools. Our estimates reveal that contributions per pupil do decline with size; however, the rate of decline is surprisingly slow. © 2002 Elsevier B.V. All rights reserved.","Brunner, E.; Sonstelie, J.",J. Public Econ.,not included
+When Should the Ask Be a Nudge? The Effect of Default Amounts on Charitable Donations,,"Goswami, I.; Urminsky, O.",Journal of Marketing Research,not included
+"Intra-organizational volunteerism: Good soldiers, good deeds and good politics","Despite the millions of hours donated to charity each year by employees on behalf of their employers there has been relatively little research into the motives for such pro-social behavior. The current paper extends Peterson's (2004, Journal of Business Ethics 49, 371) study by exploring a unique form of employee volunteerism identified as intra-organizational, or employer-sanctioned volunteerism, and uniting the heretofore distinct charity support and organizational citizenship behavior literatures. Results of a preliminary study revealed that employee participation in such intra-organizational volunteer programs is motivated by charity, firm, and personal benefits. Managerial and research implications are presented. © Springer 2006.","Peloza, J.; Hassay, D.N.",J. Bus. Ethics,not included
+Implicit vs. Explicit deception in ultimatum games with incomplete information,"We explore bargaining, using ultimatum games, when one party, the proposer, possesses private information about the pie size and can either misrepresent this information through untruthful statements (explicit deception) or through information-revealing actions (implicit deception). Our study is the first such direct comparison between two ways in which people can deceive. We find that requiring informed parties to make an explicit statement yields greater deception than when information is communicated implicitly, particularly for larger stakes. However, allowing the explicit statement to be accompanied by a promise of truthfulness reverses this effect. In contrast with many previous studies, we generally observe very high frequencies of dishonesty. © 2013 Elsevier B.V.","Kriss, P.H.; Nagel, R.; Weber, R.A.",J. Econ. Behav. Organ.,not included
+"Why people choose teaching: A scoping review of empirical studies, 2007–2016","Who enters teaching and why are questions of immense social and political importance throughout the world. This paper presents a scoping review of empirical studies, published between 2007 and 2016, that addressed influences on the choice of teaching as a career. Seventy articles were analysed descriptively and substantively. Our overview of the nature, extent, and range of research published in these articles highlights that most studies focus on motivations for teaching, with intrinsic and altruistic motivations most commonly identified. We argue that a broader range of theoretical perspectives could add fresh insights to the question of why people choose teaching. (PsycINFO Database Record (c) 2018 APA, all rights reserved)","Fray, Leanne; Gore, Jennifer",Teaching and Teacher Education,not included
+Persuasion: Theory & Research,,"O’Keefe, D.J.",Persuasion: Theory and research,not included
+Being sticker rich: Numerical context influences children's sharing behavior,"Young children spontaneously share resources with anonymous recipients, but little is known about the specific circumstances that promote or hinder these prosocial tendencies. Children (ages 3-11) received a small (12) or large (30) number of stickers, and were then given the opportunity to share their windfall with either one or multiple anonymous recipients (Dictator Game). Whether a child chose to share or not varied as a function of age, but was uninfluenced by numerical context. Moreover, children's giving was consistent with a proportion- based account, such that children typically donated a similar proportion (but different absolute number) of the resources given to them, regardless of whether they originally received a small or large windfall. The proportion of resources donated, however, did vary based on the number of recipients with whom they were allowed to share, such that on average, children shared more when there were more recipients available, particularly when they had more resources, suggesting they take others into consideration when making prosocial decisions. Finally, results indicated that a child's gender also predicted sharing behavior, with males generally sharing more resources than females. Together, findings suggest that the numerical contexts under which children are asked to share, as well as the quantity of resources that they have to share, may interact to promote (or hinder) altruistic behaviors throughout childhood. © 2015 Posid et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.","Posid, T.; Fazio, A.; Cordes, S.",PLoS ONE,not included
+What's in a message? The longitudinal influence of a supportive versus combative orientation on the performance of nonprofits,,"Botner, K.A.; Mishra, A.; Mishra, H.",Journal of Marketing Research,not included
+Advancing Measurement and Research on Youths’ Prosocial Behavior in the Digital Age,"Widespread access to digital and social media has drastically altered the nature of youth’s interpersonal connections. In this context, the opportunities children and adolescents have to help people around them are rapidly evolving. In this article, we review emerging literature on how digital media influences youth’s prosocial development in new ways. Then we propose the next steps for advancing the field’s understanding of youth’s prosocial behavior in the digital age. We advocate for extending existing measures to capture experiences that are increasingly relevant for children and adolescents today, with a focus on current events, including the COVID-19 pandemic, and social and political activism. We also provide a research agenda to advance the understanding of prosocial development. © 2021 The Authors Child Development Perspectives © 2021 The Society for Research in Child Development","Armstrong-Carter, E.; Telzer, E.H.",Child Dev. Perspect.,not included
+The cost-effectiveness of public postsecondary education subsidies,,"Muennig P, Fahs M",,not included
+Gossip as an alternative for direct observation in games of indirect reciprocity,"Communication about social topics is abundant in human societies, and many functions have been attributed to such gossiping. One of these proposed functions is the management of reputations. Reputation by itself has been shown to have a strong influence on cooperation dynamics in games of indirect reciprocity, and this notion helps to explain the observed high level of cooperation in humans. Here we designed a game to test a widespread assumption that gossip functions as a vector for the transmission of social information. This empirical study (with 14 groups of nine students each) focuses on the composition of gossip, information transfer by gossip, and the behavior based on gossip information. We show that gossip has a strong influence on the resulting behavior even when participants have access to the original information (i.e., direct observation) as well as gossip about the same information. Thus, it is evident that gossip has a strong manipulative potential. Furthermore, gossip about cooperative individuals is more positive than gossip about uncooperative individuals, gossip comments transmit social information successfully, and cooperation levels are higher when people encounter positive compared with negative gossip. © 2007 by The National Academy of Sciences of the USA.","Sommerfeld, R.D.; Krambeck, H.-J.; Semmann, D.; Milinski, M.",Proc. Natl. Acad. Sci. U. S. A.,not included
+Time to loss of brain function and activity during circulatory arrest,"PURPOSE: Brain function during the dying process and around the time of cardiac arrest is poorly understood. To better inform the clinical physiology of the dying process and organ donation practices, we performed a scoping review of the literature to assess time to loss of brain function and activity after circulatory arrest. MATERIALS AND METHODS: Medline and Embase databases were searched from inception to June 2014 for articles reporting the time interval to loss of brain function or activity after loss of systemic circulation. RESULTS: Thirty-nine studies met selection criteria. Seven human studies and 10 animal studies reported that electroencephalography (EEG) activity is lost less than 30seconds after abrupt circulatory arrest. In the setting of existing brain injury, with progressive loss of oxygenated circulation, loss of EEG may occur before circulatory arrest. Cortical evoked potentials may persist for several minutes after loss of circulation. CONCLUSION: The time required to lose brain function varied according to clinical context and method by which this function is measured. Most studies show that clinical loss of consciousness and loss of EEG activity occur within 30seconds after abrupt circulatory arrest and may occur before circulatory arrest after progressive hypoxia-ischemia. Prospective clinical studies are required to confirm these observations.","Pana, R; Hornby, L; Shemie, S D; Dhanani, S; Teitelbaum, J",J. Crit. Care,not included
+Stability of hemoglobin mass over 100 days in active men,"The purpose of this study was to investigate the suggestion in a recent meta-analysis that variability in hemoglobin mass increases when time between measurements increases from days to months. Hemoglobin mass of six active men was measured with the carbon monoxide method every 1-6 days for 100-114 days (42 +/- 3 measurements, mean +/- SD). Measurement error for each individual's series was estimated from the standard deviation of consecutive pairwise changes and compared with his total error (standard deviation of all values). Linear trends and periodicities in each series were quantified by regression and spectral analysis. Series with known random error and periodicity were also simulated and analyzed. There were clear differences in the pairwise error of measurement between subjects (range 1.4-2.7%). For five men, there was little difference between the total and pairwise errors; their mean ratio (1.06, 90% confidence limits 0.96-1.17) was less than ratios for simulated sinusoidal series with random error of 2%, amplitude of 2%, and periods of 20-100 days (ratios 1.13-1.21). Spectral analysis clearly revealed such periodicities in the simulated series but not in the series of these subjects. The sixth man, who had donated blood 12 days before commencing measurements, showed errors, trend, and periodicity consistent with gradual restoration of hemoglobin mass. Measurement error of hemoglobin mass does not increase over 100 days. Consequently, hemoglobin mass may be suitable for long-term monitoring of small changes that might occur with training or erythropoietin abuse, taking into consideration the small differences between athletes in errors and trends.","Eastwood, Annette; Hopkins, Will G; Bourdon, Pitre C; Withers, Robert T; Gore, Christopher J",J. Appl. Physiol.,not included
+The life you save may be your own,,"Schelling, T.C.",Problems in Public Expenditure Analysis,not included
+"A meta-analysis of prosocial media on prosocial behavior, aggression, and empathic concern: A multidimensional approach","Studies examining the effects of exposure to prosocial media on positive outcomes are increasing in number and strength. However, existing meta-analyses use a broad definition of prosocial media that does not recognize the multidimensionality of prosocial behavior. The aim of the current study is to conduct a meta-analysis on the effects of exposure to prosocial media on prosocial behavior, aggression, and empathic concern while examining multiple moderators that the prosocial behavior literature suggests are important to our understanding of why individuals voluntarily help others (e.g., target, type, cost). Results from 72 studies involving 243 effect sizes revealed that exposure to prosocial media was related to higher levels of prosocial behavior and empathic concern and lower levels of aggressive behavior. Moderation analyses suggest that several moderators accounted for heterogeneity in the model, including age of participant, region, media type (active vs. passive), and study design. In terms of multidimensional moderators, prosocial media had stronger effects on prosocial behavior toward strangers than did any other target and on helping and prosocial thinking but not donating or volunteering. Comparisons with other meta-analyses on media effects are made and implications for parents, media producers, and researchers are discussed. (PsycINFO Database Record","Coyne, Sarah M; Padilla-Walker, Laura M; Holmgren, Hailey G; Davis, Emilie J; Collier, Kevin M; Memmott-Elison, Madison K; Hawkins, Alan J",Dev. Psychol.,included
+Treatment strategies for osteoarthritis patients with pain and hypertension,"Out of 100 patients with osteoarthritis (OA), almost 40 have a concomitant diagnosis of hypertension. Nonsteroidal anti-inflammatory drugs (NSAIDs) and cyclooxygenase-2 (COX-2) inhibitors may trigger a rise in blood pressure (BP), which is more marked in patients with established hypertension. NSAIDs and COX-2 inhibitors attenuate the antihypertensive effect of several antihypertensive agents. Frequent BP controls are needed in treated hypertensive patients who are concomitantly receiving NSAIDs or COX-2 inhibitors because even a small increase in BP may be associated with an important rise in the risk of major cardiovascular complications. In meta-analyses, an increase in systolic BP of 5mmHg was associated with a 25% higher risk of cardiovascular events. These data have been confirmed in randomized studies with rofecoxib and celecoxib, where a modest increase in BP was associated with a significantly higher risk of cardiovascular disease. There is emerging evidence that the COX-inhibiting nitric oxide donator (CINOD) class is promising in the treatment of patients with OA. Naproxcinod, the first CINOD investigated in clinical trials, is composed of the traditional NSAID naproxen covalently bound to the nitric oxide (NO)-donating moiety butanediol mono-nitrate (BDMN). The molecule has the potential to provide a sustained release of NO. In clinical studies, naproxcinod prevented the BP rise in normotensive and hypertensive patients observed with naproxen. The BP benefit of naproxcinod over naproxen was greater in patients concomitantly receiving angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers. These investigational data suggest that naproxcinod is a valuable alternative to NSAIDs and COX-2 inhibitors for treatment of OA patients.","Verdecchia, Paolo; Angeli, Fabio; Mazzotta, Giovanni; Martire, Paola; Garofoli, Marta; Gentile, Giorgio; Reboldi, Gianpaolo",Ther. Adv. Musculoskelet. Dis.,not included
+"Nonprofit organizations, monopolistic competition, and private donations: Evidence from Spain","This article presents an analysis of the determinants of money and time donations to Spanish nongovernmental organizations that channel aid to less developed countries. A basic model inspired by the theory of monopolistic competition is formulated and tested taking into account that some of the explanatory variables, such as fund-raising expenditure and price, are endogenous. The results show that the average donor is different for money and time donations and that government preferences differ from those of private donors. Finally, the authors find that the hypothesis of efficient fund-raising expenditures cannot be rejected.","Marcuello, C.; Salas, V.",Public Financ. Rev.,not included
+Public goods provision and redistributive taxation,"This paper studies the relationship between redistributive taxation and tax-deductible charitable contributions. Redistribution has two opposite effects on voluntary giving. The price of charitable giving decreases with the degree of redistribution, and this has a positive effect on the total amount of giving (substitution effect). However, redistribution leads to lower consumption for the contributors and therefore has a negative effect on contributions to the charity (income effect). The theoretical model developed in this paper demonstrates that, under a general class of utility functions, the substitution effect dominates the income effect. Hence, charitable giving increases with the tax rate. In purely egalitarian societies, the public good is provided efficiently and the total welfare is maximized independent of the ex-ante income inequality. However, the positive impact of taxation on charitable giving and welfare may disappear if individuals generate their income levels in anticipation of taxation and redistribution does not take into account the cost of effort. © 2008 Elsevier B.V. All rights reserved.","Uler, N.",J. Public Econ.,not included
+"Responsibility, norms, and helping in an emergency","Replicated the J. Darley and B. Latane (see record) study of bystander aid to a seizure victim examining the effects of (a) number and competence of bystanders, (b) information appropriate for action, and (c) ascription of responsibility (AR) upon helping by males and females. From an analysis of norms relevant in an emergency and of the likelihood of their activation, main effects on speed of helping for the above 4 variables, interactions of the 1st 3 with AR, a Sex of Subject × Number interaction, and differences in type of help offered in various conditions were predicted. 179 undergraduates participated in a factorial experiment. Speed of helping dropped significantly for females, but not for males, when other bystanders were present (reporting decreased, direct help was unaffected), and dropped significantly further when another bystander was medically competent (reporting increased, direct help decreased). Among females disposed to accept rationales for denying responsibility, both effects were particularly strong. Information-action and AR to the self were associated with faster and more direct help. Data on Ss' thoughts and feelings reinforced a normative interpretation of the results. (PsycINFO Database Record (c) 2006 APA, all rights reserved). © 1970 American Psychological Association.","Schwartz, S.H.; Clausen, G.T.",J. Pers. Soc. Psychol.,not included
+Emerging adulthood: A theory of development from the late teens through the twenties,"Emerging adulthood is proposed as a new conception of development for the period from the late teens through the twenties, with a focus on ages 18-25. A theoretical background is presented. Then evidence is provided to support the idea that emerging adulthood is a distinct period demographically, subjectively, and in terms of identity explorations. How emerging adulthood differs from adolescence and young adulthood is explained. Finally, a cultural context for the idea of emerging adulthood is outlined, and it is specified that emerging adulthood exists only in cultures that allow young people a prolonged period of independent role exploration during the late teens and twenties.","Arnett, J.J.",Am. Psychol.,not included
+Assessing actual strategic behavior to construct a measure of strategic ability,"Strategic interactions have been studied extensively in the area of judgment and decision-making. However, so far no specific measure of a decision-maker's ability to be successful in strategic interactions has been proposed and tested. Our contribution is the development of a measure of strategic ability that borrows from both game theory and psychology. Such measure is aimed at providing an estimation of the likelihood of success in many social activities that involve strategic interaction among multiple decision-makers. To construct a reliable measure of strategic ability, that we propose to call ""Strategic Quotient"" (SQ), we designed a test where each item is a game and where, therefore, the individual obtained score depends on the distribution of choices of other decision-makers taking the test. The test is designed to provide information on the abilities related to two dimensions, mentalization and rationality, that we argue are crucial to strategic success, with each dimension being characterized by two main factors. Principal component analysis on preliminary data shows that indeed four factors (two for rationality, two for mentalization) account for strategic success in most of the strategically simpler games of the test. Moreover, two more strategically sophisticated games are inserted in the test and are used to investigate if and to what extent the four factors obtained by simpler games can predict strategic success in more sophisticated strategic interactions. Overall, the collected empirical evidence points to the possibility of building a SQ measure using only simple games designed to capture information about the four identified factors. © 2019 Bilancini, Boncinelli and Mattiassi.","Bilancini, E.; Boncinelli, L.; Mattiassi, A.",Front. Psychol.,not included
+Impact of presumed consent for organ donation on donation rates: a systematic review,"OBJECTIVES: To examine the impact of a system of presumed consent for organ donation on donation rates and to review data on attitudes towards presumed consent. DESIGN: Systematic review. DATA SOURCES: Studies retrieved by online searches to January 2008 of Medline, Medline In-Process, Embase, CINAHL, PsycINFO, HMIC, PAIS International, and OpenSIGLE. Studies reviewed Five studies comparing donation rates before and after the introduction of legislation for presumed consent (before and after studies); eight studies comparing donation rates in countries with and without presumed consent systems (between country comparisons); 13 surveys of public and professional attitudes to presumed consent. RESULTS: The five before and after studies represented three countries: all reported an increase in donation rates after the introduction of presumed consent, but there was little investigation of any other changes taking place concurrently with the change in legislation. In the four best quality between country comparisons, presumed consent law or practice was associated with increased organ donation-increases of 25-30%, 21-26%, 2.7 more donors per million population, and 6.14 more donors per million population in the four studies. Other factors found to be important in at least one study were mortality from road traffic accidents and cerebrovascular causes, transplant capacity, gross domestic product per capita, health expenditure per capita, religion (Catholicism), education, public access to information, and a common law legal system. Eight surveys of attitudes to presumed consent were of the UK public. These surveys varied in the level of support for presumed consent, with surveys conducted before 2000 reporting the lowest levels of support (28-57%). The most recent survey, in 2007, reported that 64% of respondents supported a change to presumed consent. CONCLUSION: Presumed consent alone is unlikely to explain the variation in organ donation rates between countries. Legislation, availability of donors, organisation and infrastructure of the transplantation service, wealth and investment in health care, and public attitudes to and awareness of organ donation may all play a part, but their relative importance is unclear. Recent UK surveys show support for presumed consent, though with variation in results that may reflect differences in survey methods.","Rithalia, Amber; McDaid, Catriona; Suekarran, Sara; Myers, Lindsey; Sowden, Amanda",BMJ,not included
+Social learning theory,"Three forms of social learning theory relevant to child development are reviewed. The first (see Robert Sears) involved attempts to combine Freudian and stimulus-response learning theory. Explaining the internalization of parent expectations, however, was a challenge for this approach. In a second form, Albert Bandura employed learning principles of reinforcement and punishment but also argued that the primary form of learning was observation. Additionally, he emphasized the role of cognition in learning. A third form, associated with Gerald Patterson among others, focuses on behavior management of difficult children and continues to underlie many forms of intervention practiced today. © 2019 Elsevier Inc. All rights reserved.","Grusec, J.E.",The Curated Reference Collection in Neuroscience and Biobehavioral Psychology,not included
+Social dilemma cooperation (unlike Dictator Game giving) is intuitive for men as well as women,"Does intuition favor prosociality, or does prosocial behavior require deliberative self-control? The Social Heuristics Hypothesis (SHH) stipulates that intuition favors typically advantageous behavior - but which behavior is typically advantageous depends on both the individual and the context. For example, non-zero-sum cooperation (e.g. in social dilemmas like the Prisoner's Dilemma) typically pays off because of the opportunity for reciprocity. Conversely, reciprocity does not promote zero-sum cash transfers (e.g. in the Dictator Game, DG). Instead, DG giving can be long-run advantageous because of reputation concerns: social norms often require such behavior of women but not men. Thus, the SHH predicts that intuition will favor social dilemma cooperation regardless of gender, but only favor DG giving among women. Here I present meta-analytic evidence in support of this prediction. In 31 studies examining social dilemma cooperation (N=13,447), I find that promoting intuition increases cooperation to a similar extent for both men and women. This stands in contrast to the results from 22 DG studies (analyzed in Rand et al., 2016) where intuition promotes giving among women but not men. Furthermore, I show using meta-regression that the interaction between gender and intuition is significantly larger in the DG compared to the cooperation games. Thus, I find clear evidence that the role of intuition and deliberation varies across both setting and individual as predicted by the SHH.","Rand, David G",J. Exp. Soc. Psychol.,not included
+Public charity offer as a proximate factor of evolved reputation-building strategy: an experimental analysis of a real-life situation,"Although theoretical considerations suggest that a considerable portion of human altruism is driven by concerns about reputation, few experimental studies have examined the psychological correlates of individual decisions in real-life situations. Here we demonstrate that more subjects were willing to give assistance to unfamiliar people in need if they could make their charity offers in the presence of their group mates than in a situation where the offers remained concealed from others. In return, those who were willing to participate in a particular charitable activity received significantly higher scores than others on scales measuring sympathy and trustworthiness. Finally, a multiple regression analysis revealed that while several personality and behavior traits (cooperative ability, Machiavellianism, sensitivity to norms, and sex) play a role in the development of prosocial behavior, the possibility of gaining reputation within the group remains a measurable determinant of charitable behavior. © 2007 Elsevier Inc. All rights reserved.","Bereczkei, T.; Birkas, B.; Kerekes, Z.",Evol. Hum. Behav.,not included
+A comparison of two behavioral influence techniques for improving blood donor recruitment,"This study was designed to test the viability of two multiple request techniques of behavioral influence for recruiting blood donors by telephone. The first technique utilizes a small antecedent request to encourage behavioral involvement and favorable disposition toward the target activity of the critical request to donate. The second approach frames the critical request as a concession following refusal of a very large request. The two techniques, dubbed the foot‐in‐the‐door (FID) and door‐in‐the‐face (DIF), respectively, were tested against a control condition on three donor groups: active donors, inactive donors, and nondonors. Thus, a three‐by‐three factorial design was used on 910 adults in a Midwest city. Although the DIF was outperformed by the control across all three donor groups, the authors recommend its continued study in face‐to‐face donor solicitation. Importantly, the FID approach produced more donations than the control condition among active donors (Z = 4.30; p < .001), inactives (Z = 7.45; p < .001), and nondonors (Z = 1.98; p < .05). For managing the blood supply, the FID is particularly potent for rekindling donations from inactive donors. Additional research on means of penetrating the nondonor segment is recommended. 1984 AABB","Dwyer, F.R.; Greenwalt, T.J.; Coe, N.A.",Transfusion,not included
+Adult‐related haematopoietic stem cell donor experiences and the provision of information and psychosocial support: A systematic literature review,"For blood cancer patients, haematopoietic stem cells (HSC) donated by a relative can be lifesaving. However, related donors can face significant physical and psychosocial challenges. As the demand for adult‐related HSC donors is increasing, it is important to review our understanding of adult‐related HSC donors’ need for and availability of information and psychosocial support with a view to identifying gaps in the literature. A systematic review of relevant studies (2000–2017) was conducted using five databases with supplementary hand searching. Sixteen studies involving 1,024 related HSC donors met the following criteria: English or Dutch language, peer‐reviewed, sampled first‐time‐related HSC donors, ≥ 18 years, haematological malignancies, assessed psychosocial aspects, retrospective or prospective and with or without comparison group. Data were abstracted, and study quality was assessed using the PRISMA criteria. Studies contained limited information on the provision of information and psychosocial support. Most studies addressed pre‐donation information, and none reported providing information or support to donors post‐donation. Additionally, few studies formally assessed unmet needs. Recommendations include improved transparency of reporting for the availability, sources and timing of information and psychosocial support, and the identification of unmet needs to enable the development of educational and psychosocial interventions for this invaluable donor population.patients, haematopoietic stem cells (HSC) donated by a relative can be lifesaving. However, related donors can face significant physical and psychosocial challenges. As the demand for adult‐related HSC donors is increasing, it is important to review our understanding of adult‐related HSC donors’ need for and availability of information and psychosocial support with a view to identifying gaps in the literature. A systematic review of relevant studies (2000 (PsycINFO Database Record (c) 2019 APA, all rights reserved)","Zomerdijk, Nienke; Turner, Jane M; Hill, Geoffrey R",Eur. J. Cancer Care,not included
+Scaling up the 2010 World Health Organization HIV treatment guidelines in resource-limited settings: a model-based analysis,,"Walensky, R P; Wood, R; Ciaranello, A L; Paltiel, A D; Lorenzana, S B; Anglaret, X; Stoler, A W; Freedberg, K A; Cost Effectiveness of AIDS Complications International Investigators",,not included
+"""Paper or plastic?"": How we pay influences post-transaction connection",,"Shah, A.M.; Eisenkraft, N.; Bettman, J.R.; Chartrand, T.L.",Journal of Consumer Research,not included
+Meta-analysis for public management & policy,,"Ringquist, E.",Meta-Analysis for Public Management and Policy,not included
+Effects of sexual dimorphism on facial attractiveness,"Testosterone-dependent secondary sexual characteristics in males may signal immunological competence and are sexually selected for in several species. In humans, oestrogen-dependent characteristics of the female body correlate with health and reproductive fitness and are found attractive. Enhancing the sexual dimorphism of human faces should raise attractiveness by enhancing sex-hormone-related cues to youth and fertility in females, and to dominance and immunocompetence in males. Here we report the results of asking subjects to choose the most attractive faces from continua that enhanced or diminished differences between the average shape of female and male faces. As predicted, subjects preferred feminized to average shapes of a female face. This preference applied across UK and Japanese populations but was stronger for within-population judgements, which indicates that attractiveness cues are learned. Subjects preferred feminized to average or masculinized shapes of a male face. Enhancing masculine facial characteristics increased both perceived dominance and negative attributions (for example, coldness or dishonesty) relevant to relationships and paternal investment. These results indicate a selection pressure that limits sexual dimorphism and encourages neoteny in humans.","Perrett, D.I.; Lee, K.J.; Penton-Voak, I.; Rowland, D.; Yoshikawa, S.; Burt, D.M.; Henzi, S.P.; Castles, D.L.; Akamatsu, S.",Nature,not included
+"The cost-effectiveness of introducing nucleic acid testing to test for hepatitis B, hepatitis C, and human immunodeficiency virus among blood donors in Sweden",,"Davidson, T; Ekermo, B; Gaines, H; Lesko, B; Akerlind, B",,not included
+Does government funding suppress nonprofits' political activity?,"Autonomy from the state has been considered a core feature of American civil society, and understanding the consequences of perceived threats to that autonomy has been a central theme in social and political theory. We engage this theme by examining a specific question: What is the effect of government funding on nonprofit organizations' political activity? Extant theory and research identify some mechanisms by which government funding might reduce nonprofit political activity and other mechanisms by which government funding might enhance such activity. We investigate this relationship with two data sets: a national sample of religious congregations and a longitudinal sample of nonprofit organizations in Minneapolis-St. Paul. Results across these data sets are consistent and compelling: The relationship between government funding and nonprofit political activity is either positive or null; government funding does not suppress nonprofit political activity.","Chaves, M.; Stephens, L.; Galaskiewicz, J.",Am. Sociol. Rev.,not included
+Governing the Hollow State,"For the past ten years the authors have conducted a concentrated research program on the dimensions and impact of the hollow state. The hollow state is a metaphor for the increasing use of third parties, often nonprofits, to deliver social services and generally act in the name of the state. The types of structures, incentives, and mechanisms used to control third-party providers have been the focus of this research. The empirical thrust of this research is on how effective various types of mechanisms, structures, and incentives are at promoting the effectiveness of contracted services. The normative question this research has raised, but not answered, is, What effect does government contracting with third-party providers have on the perceived legitimacy of the state? This article is a summary of the theoretical development and the empirical findings from the authors' research on the dimensions and impact of the hollow state in the domain of health and human services contracting. Elements of this article have appeared previously in this journal and in many others as well. The article's purpose is to integrate the authors' research on the hollow state. This is a summative article that seeks to bring together in one place what the authors have learned. In addition, new directions are explored for future research on the hollow state.","Milward, H.B.; Provan, K.G.",J. Public Adm. Res. Theory,not included
+The Role of Food Banks in Addressing Food Insecurity: A Systematic Review,"Food banks play a major role in the food aid sector by distributing donated and purchased groceries directly to food insecure families. The public health implications of food insecurity are significant, particularly as food insecurity has a higher prevalence among certain population groups. This review consolidates current knowledge about the function and efficacy of food banks to address food insecurity. A systematic review was conducted. Thirty-five publications were reviewed, of which 14 examined food security status, 13 analysed nutritional quality of food provided, and 24 considered clients' needs in relation to food bank use. This review found that while food banks have an important role to play in providing immediate solutions to severe food deprivation, they are limited in their capacity to improve overall food security outcomes due to the limited provision of nutrient-dense foods in insufficient amounts, especially from dairy, vegetables and fruits. Food banks have the potential to improve food security outcomes when operational resources are adequate, provisions of perishable food groups are available, and client needs are identified and addressed.","Bazerghi, Chantelle; McKay, Fiona H; Dunn, Matthew",J. Community Health,not included
+Thinking about fit and donation format in cause marketing: The effects of need for cognition,,"Kerr, A.; Das, N.",Journal of Marketing Theory and Practice,not included
+The 'I' of the beholder: How gender differences and self-referencing influence charity advertising,,"Chang, C.-T.; Lee, Y.-K.",International Journal of Advertising,not included
+Operating characteristics of a rank correlation test for publication bias,"An adjusted rank correlation test is proposed as a technique for identifying publication bias in a meta-analysis, and its operating characteristics are evaluated via simulations. The test statistic is a direct statistical analogue of the popular 'funnel-graph.' The number of component studies in the meta-analysis, the nature of the selection mechanism, the range of variances of the effect size estimates, and the true underlying effect size are all observed to be influential in determining the power of the test. The test is fairly powerful for large meta-analyses with 75 component studies, but has only moderate power for meta-analyses with 25 component studies. However, in many of the configurations in which there is low power, there is also relatively little bias in the summary effect size estimate. Nonetheless, the test must be interpreted with caution in small meta-analyses. In particular, bias cannot be ruled out if the test is not significant. The proposed technique has potential utility as an exploratory tool for meta-analysts, as a formal procedure to complement the funnel- graph.","Begg, C.B.; Mazumdar, M.",BIOMETRICS,not included
+When will price increases associated with company donations to charity be perceived as fair?,,"Koschate-Fischer, N.; Huber (née Stefan), I.V.; Hoyer, W.D.",Journal of the Academy of Marketing Science,not included
+The Effects of Monetary Incentives and Labeling on the Foot-in-the-Door Effect: Evidence for a Self-Perception Process,"We tested the self-perception explanation of the foot-in-the-door effect by manipulating self-perceived helpfulness and assessing self-concept. Participants given $1 to sign a homelessness petition were less likely to see themselves as altruistic than participants not given the monetary incentive. The paid participants also complied less often with a request to work on a canned food drive 2 days later than unpaid participants. In contrast, participants told they were helpful individuals were more likely to see themselves as altruistic and were more likely to volunteer for the food drive than unlabeled participants. Mediation analyses provide evidence that changes in self-concept underlie a successful foot-in-the-door manipulation and support the self-perception explanation for the foot-in-the-door effect.","Burger, Jerry M; Caldwell, David F",Basic Appl. Soc. Psych.,not included
+A comparison of clinical officers with medical doctors on outcomes of caesarean section in the developing world: meta-analysis of controlled studies,"The authors tentatively concluded that there was no statistically significant difference in maternal or perinatal mortality in caesarean sections carried out by clinical officers compared with doctors, but wound dehiscence and wound infection were significantly more frequent in caesarean sections carried out by clinical officers. XCM: The review question was clear and supported by potentially reproducible inclusion criteria. The search strategy appeared to include a number of relevant sources and was not restricted by language, which reduced the possibility of language bias. It did not appear that specific searches were undertaken for unpublished studies, so some potentially relevant data may have been missed. Study selection was conducted in duplicate, but it was unclear whether similar methods to reduce error and bias were used for quality assessment and data extraction.Study quality was assessed using an appropriate tool and results were reported. Adequate details of primary studies were provided. Combining the results in meta-analyses may not have been appropriate given the variability of the studies and the statistical (and clinical) heterogeneity in the some of the meta-analyses.The authors' conclusions reflect the evidence presented, but given the variability between (and methodological shortcomings within) the included studies, together with poor reporting of the review process, their reliability is uncertain. XIM: Practice: The authors stated that there may be a particular training need for clinical officers in light of the increase in wound infection and dehiscence compared with doctors.Research: The authors did not state any implications for further research.","Wilson, A; Lissauer, D; Thangaratinam, S; Khan, K S; MacArthur, C; Coomarasamy, A",,not included
+A systematic review of episodic volunteering in public health and other contexts,"BACKGROUND: Episodic volunteers are a critical resource for public health non-profit activities but are poorly understood. A systematic review was conducted to describe the empirical evidence about episodic volunteering (EV) in the public health sector and more broadly. Study location, focus and temporal trends of EV research were also examined. METHODS: Twelve key bibliographic databases (1990-April week 2, 2014) were searched, including Google Scholar. Empirical studies published in English in peer-reviewed journals that identified participants as EVs who volunteered to support Not-for-Profit organisations in the health and social welfare sectors were included. EV definitions, characteristics, economic costs, antecedents and outcomes and theoretical approaches were examined. RESULTS: 41 articles met initial review criteria and 20 were specific to the health or social welfare sectors. EV definitions were based on one or more of three dimensions of duration, frequency, and task. EVs were predominantly female, middle aged, Caucasian (North American) and college/university educated. Fundraising was the most common EV activity and 72% had volunteered at least once. No studies examined the economic costs of EV. There was little consistency in EV antecedents and outcomes, except motives which primarily related to helping others, forming social connections, and self-psychological or physical enhancement. Most studies were atheoretical. Three authors proposed new theoretical frameworks. CONCLUSIONS: Research is required to underpin the development of an agreed consensus definition of EV. Moreover, an EV evidence-base including salient theories and measures is needed to develop EV engagement and retention strategies for the health and social welfare sectors.","Hyde, Melissa K; Dunn, Jeff; Scuffham, Paul A; Chambers, Suzanne K",BMC Public Health,not included
+Imagine being a nice guy: A note on hypothetical vs. Incentivized social preferences,"We conducted an experimental study on social preferences using dictator games similar to Fehr et al. (2008). Our results show that social preferences differ between subjects who receive low-stakes monetary rewards for their decisions and subjects who consider hypothetical stakes. Our findings indicate that, apart from incentives, gender plays an important role for the categorization of different social preferences. © 2015. The authors license.","Bühren, C.; Kundt, T.C.",Judgm. Decis. Mak.,not included
+Environmental certification programs: How does information provision compare with taxation?,"This paper develops a monopolistic competition framework to assess whether environmental certification programs can serve as effective substitutes for more traditional policy instruments such as environmental taxation or a minimum quality standard (MQS). I show that if firms can organize themselves and choose the certification standard collectively, then there is a beneficial role for a regulator to intervene. Also, the degree of substitution between differentiated goods that impose environmental damage and a “clean” outside good, the degree of competition in the industry and the extent of environmental damage caused by minimal quality goods are important considerations in the choice between a certification program and a tax or a MQS. While the comparison between a certification program and a tax depends on numerous factors, I find unequivocally that certification is a poor substitute for taxation whenever the outside good is a close substitute for differentiated goods, there is a high degree of competition in the industry or if minimal quality goods impose considerable environmental damage. © 2020 Wiley Periodicals LLC","Podhorsky, A.",J. Public Econ. Theory,not included
+A website to host educational modules on global engineering ethics and conduct research in cross-cultural moral psychology: A work in progress,"To ensure more long-term ethical behaviors within engineering, a website is being developed to host educational modules on global engineering ethics and conduct research on cross-cultural moral psychology. The modules are all-inclusive, with a cross-cultural and international focus, requiring less preparation on the part of instructors and are easier for different types of students to use than existing online resources. Education and research using the site can occur at the same time, each strengthening the other in the process. Rather than simply ethical understanding or the ability to reason ethically, research on moral psychology can ensure more ethical behaviors, better understanding what people know and think about ethics and the causes of (un)ethical behaviors. This research is cross-cultural, since culture has been shown to affect behaviors and thoughts related to ethics, and the educational and working environments of engineering are more cross-cultural and international than ever before. © American Society for Engineering Education, 2019","Clancy, R.F., III; Manuel, C.",ASEE Annu. Conf. Expos. Conf. Proc.,not included
+Influence of various models on aggressive behavior in individuals with different socialization experience,,"Toeplitz-Winiewska, M.",Polish Psychological Bulletin,not included
+Consumer reaction to price increase: An investigation in gasoline industry,"Purpose – The aim of this study is to investigate the impact of increase in price of an essential product (i.e. gasoline) toward the focal product and other seemingly non-related products. Design/methodology/approach – A self-administered survey was used to collect data from the drivers at a large metroplex in Southwest USA. Multiple regression and scanning electron microscope procedures were used to analyze and test the proposed hypotheses. Findings – When consumers notice the increase in gas prices, they become very anxious. This anxiety is positively associated with average gas bought in gallons and negatively associated with threshold price. Further, this consumer anxiety has the strongest influence on lifestyle changes, followed by automobile technology change and transportation mode change, and has the weakest influence on gasoline brand/type change. Research limitations/implications – We focus on only anxiety as a mediator between increase in gas prices and the behavioral outcomes, and collect data from only one location. Practical implications – Managers must be cognizant that a price increase in essential goods not only influences the demand for focal products but also for products that may not seem related to the focal products. Social implications – Increase in gasoline price will not only affect the demand for gasoline, but also the demand for alternate forms of transportation, fuel efficient vehicles, and other aspects of life. Originality/value – This study is the first to look at the role of anxiety as a mediator and looks at the effects of increase in gas prices in a holistic manner. © Emerald Group Publishing Limited.","Paswan, A.K.; Crawford, J.C.; Ngamsiriudom, W.; Nguyen, T.",J. Prod. Brand Manage.,not included
+Internally Reporting Risk in Financial Services: An Empirical Analysis,"The enduring failure of financial institutions to identify and deal with risk events continues to have serious repercussions, whether in the form of small but significant losses or major and potentially far-reaching scandals. Using a mixed-methods approach that combines an innovative version of the classic dictator game to inform prosocial tendencies with the survey-based Theory of Planned Behaviour, we examine the risk-escalation behaviour of individuals within a large financial institution. We discover evidence of purely selfish behaviour that explains the lack significance in pressure to adhere to the Subjective Norms of colleagues around intention to report risks. A finding that has potentially important implications for efforts to instil a high-error management climate and incentivise risk reporting within organisations where risk, if ignored or unchecked, could ultimately have consequences that extend far beyond the institutions themselves.","Bryce, Cormac; Chmura, Thorsten; Webb, Rob; Stiebale, Joel; Cheevers, Carly",J. Bus. Ethics,not included
+,,,"Tax Reform for Fairness, Simplicity, and Economic Growth",not included

data/tai_safety_research/task.json

@@ -0,0 +1 @@
+{"name": "tai_safety_research", "description": "", "data_columns": ["Title", "Abstract Note", "Url", "Publication Year", "Item Type", "Author", "Publication Title"], "label_columns": {"Label": ["TAI safety research", "not TAI safety research"]}}

data/tai_safety_research/train.csv

@@ -0,0 +1,51 @@
+Title,Abstract Note,Url,Publication Year,Item Type,Author,Publication Title,Label
+Malign generalization without internal search,"In my last post, I challenged the idea that inner alignment failures should be explained by appealing to agents which perform explicit internal search. By doing so, I argued that we should instead appeal to the more general concept of  malign generalization, and treat mesa-misalignment as a special case.  Unfortunately, the post was light on examples of what we should be worrying about instead of mesa-misalignment. Evan Hubinger wrote, Personally, I think there is a meaningful sense in which all the models I'm most worried about do some sort of search internally (at least to the same extent that humans do search internally), but I'm definitely uncertain about that.Wei Dai expressed confusion why I would want to retreat to malign generalization without some sort of concrete failure mode in mind, Can you give some realistic examples/scenarios of “malign generalization” that does not involve mesa optimization? I’m not sure what kind of thing you’re actually worried about here.In this post, I will outline a general category of agents which may exhibit malign generalization without internal search, and then will provide a concrete example of an agent in the category. Then I will argue that, rather than being a very narrow counterexample, this class of agents could be competitive with search-based agents.  THE SWITCH CASE AGENT Consider an agent governed by the following general behavior,  LOOP:State = GetStateOfWorld(Observation)IF State == 1:PerformActionSequence1() IF State == 2:PerformActionSequence2()...END_LOOP  It's clear that this agent does not perform any internal search for strategies: it doesn't operate by choosing actions which rank highly according to some sort of internal objective function. While you could potentially rationalize its behavior according to some observed-utility function, this would generally lead to more confusion than clarity. However, this agent could still be malign in the following way. Suppose the agent is 'mistaken' about the s",https://www.alignmentforum.org/posts/ynt9TD6PrYw6iT49m/malign-generalization-without-internal-search,2020,blogPost,"Barnett, Matthew",AI Alignment Forum,TAI safety research
+Utility Indifference,"Consider an AI that follows its own motivations. We’re not entirely sure what its motivations are, but we would prefer that the AI cooperate with humanity; or, failing that, that we can destroy it before it defects. We’ll have someone sitting in a room, their finger on a detonator, ready at the slightest hint of defection. Unfortunately as has been noted ([3], [1]), this does not preclude the AI from misbehaving. It just means that the AI must act to take control of the explosives, the detonators or the human who will press the button. For a superlatively intelligence AI, this would represent merely a slight extra difficulty. But now imagine that the AI was somehow indifferent to the explosives going off or not (but that nothing else was changed). Then if ever the AI does decide to defect, it will most likely do so without taking control of the explosives, as that would be easier than otherwise. By “easier ” we mean that the chances of failure are less, since the plan is simpler – recall that under these assumptions, the AI counts getting blown up as an equal value to successfully defecting.",,2010,report,"Armstrong, Stuart",,TAI safety research
+Improving Sample Efficiency in Model-Free Reinforcement Learning from Images,"Training an agent to solve control tasks directly from high-dimensional images with model-free reinforcement learning (RL) has proven difficult. A promising approach is to learn a latent representation together with the control policy. However, fitting a high-capacity encoder using a scarce reward signal is sample inefficient and leads to poor performance. Prior work has shown that auxiliary losses, such as image reconstruction, can aid efficient representation learning. However, incorporating reconstruction loss into an off-policy learning algorithm often leads to training instability. We explore the underlying reasons and identify variational autoencoders, used by previous investigations, as the cause of the divergence. Following these findings, we propose effective techniques to improve training stability. This results in a simple approach capable of matching state-of-the-art model-free and model-based algorithms on MuJoCo control tasks. Furthermore, our approach demonstrates robustness to observational noise, surpassing existing approaches in this setting. Code, results, and videos are anonymously available at https://sites.google.com/view/sac-ae/home.",http://arxiv.org/abs/1910.01741,2020,manuscript,"Yarats, Denis; Zhang, Amy; Kostrikov, Ilya; Amos, Brandon; Pineau, Joelle; Fergus, Rob",,not TAI safety research
+Teaching A.I. Systems to Behave Themselves (Published 2017),"As philosophers and pundits worry that artificial intelligence will one day harm the world, some researchers are working on ways to lower the risks.",https://www.nytimes.com/2017/08/13/technology/artificial-intelligence-safety-training.html,2017,newspaperArticle,"Metz, Cade",The New York Times,not TAI safety research
+Incentives in Teams,,https://www.jstor.org/stable/1914085?origin=crossref,1973,journalArticle,"Groves, Theodore",Econometrica,not TAI safety research
+A bargaining-theoretic approach to moral uncertainty,"This paper explores a new approach to the problem of decision under relevant moral uncertainty. We treat the case of an agent making decisions in the face of moral uncertainty on the model of bargaining theory, as if the decision-making process were one of bargaining among different internal parts of the agent, with different parts committed to different moral theories. The resulting approach contrasts interestingly with the extant “maximise expected choiceworthiness” and “my favourite theory” approaches, in several key respects. In particular, it seems somewhat less prone than the MEC approach to ‘fanaticism’: allowing decisions to be dictated by a theory in which the agent has extremely low credence, if the relative stakes are high enough. Overall, however, we tentatively conclude that the MEC approach is superior to a bargaining-theoretic approach.",,2019,report,"Greaves, Hilary; Cotton-Barratt, Owen",,not TAI safety research
+The Timing of Evolutionary Transitions Suggests Intelligent Life Is Rare,"It is unknown how abundant extraterrestrial life is, or whether such life might be complex or intelligent. On Earth, the emergence of complex intelligent life required a preceding series of evolutionary transitions such as abiogenesis, eukaryogenesis, and the evolution of sexual reproduction, multicellularity, and intelligence itself. Some of these transitions could have been extraordinarily improbable, even in conducive environments. The emergence of intelligent life late in Earth's lifetime is thought to be evidence for a handful of rare evolutionary transitions, but the timing of other evolutionary transitions in the fossil record is yet to be analyzed in a similar framework. Using a simplified Bayesian model that combines uninformative priors and the timing of evolutionary transitions, we demonstrate that expected evolutionary transition times likely exceed the lifetime of Earth, perhaps by many orders of magnitude. Our results corroborate the original argument suggested by Brandon Carter that intelligent life in the Universe is exceptionally rare, assuming that intelligent life elsewhere requires analogous evolutionary transitions. Arriving at the opposite conclusion would require exceptionally conservative priors, evidence for much earlier transitions, multiple instances of transitions, or an alternative model that can explain why evolutionary transitions took hundreds of millions of years without appealing to rare chance events. Although the model is simple, it provides an initial basis for evaluating how varying biological assumptions and fossil record data impact the probability of evolving intelligent life, and also provides a number of testable predictions, such as that some biological paradoxes will remain unresolved and that planets orbiting M dwarf stars are uninhabitable.",https://www.liebertpub.com/doi/full/10.1089/ast.2019.2149,2020,journalArticle,"Snyder-Beattie, Andrew E.; Sandberg, Anders; Drexler, K. Eric; Bonsall, Michael B.",Astrobiology,not TAI safety research
+Changing Identity: Retiring from Unemployment,,https://academic.oup.com/ej/article/124/575/149-166/5076984,2014,journalArticle,"Hetschko, Clemens; Knabe, Andreas; Schöb, Ronnie",The Economic Journal,not TAI safety research
+Model-Based Reinforcement Learning via Meta-Policy Optimization,"Model-based reinforcement learning approaches carry the promise of being data efficient. However, due to challenges in learning dynamics models that sufficiently match the real-world dynamics, they struggle to achieve the same asymptotic performance as model-free methods. We propose Model-Based Meta-Policy-Optimization (MB-MPO), an approach that foregoes the strong reliance on accurate learned dynamics models. Using an ensemble of learned dynamic models, MB-MPO meta-learns a policy that can quickly adapt to any model in the ensemble with one policy gradient step. This steers the meta-policy towards internalizing consistent dynamics predictions among the ensemble while shifting the burden of behaving optimally w.r.t. the model discrepancies towards the adaptation step. Our experiments show that MB-MPO is more robust to model imperfections than previous model-based approaches. Finally, we demonstrate that our approach is able to match the asymptotic performance of model-free methods while requiring significantly less experience.",http://arxiv.org/abs/1809.05214,2018,manuscript,"Clavera, Ignasi; Rothfuss, Jonas; Schulman, John; Fujita, Yasuhiro; Asfour, Tamim; Abbeel, Pieter",,not TAI safety research
+Advancing rational analysis to the algorithmic level,"Abstract             The commentaries raised questions about normativity, human rationality, cognitive architectures, cognitive constraints, and the scope or resource rational analysis (RRA). We respond to these questions and clarify that RRA is a methodological advance that extends the scope of rational modeling to understanding cognitive processes, why they differ between people, why they change over time, and how they could be improved.",https://www.cambridge.org/core/product/identifier/S0140525X19002012/type/journal_article,2020,journalArticle,"Lieder, Falk; Griffiths, Thomas L.",Behavioral and Brain Sciences,not TAI safety research
+Confronting future catastrophic threats to humanity,,https://linkinghub.elsevier.com/retrieve/pii/S0016328715001135,2015,journalArticle,"Baum, Seth D.; Tonn, Bruce E.",Futures,TAI safety research
+Latent Variables and Model Mis-Specification,"Posted as part of the AI Alignment Forum sequence on Value Learning. Rohin's note: So far, we’ve seen that ambitious value learning needs to understand human biases, and that we can't simply learn the biases in tandem with the reward. Perhaps we could hardcode a specific model of human biases? Such a model is likely to be incomplete and inaccurate, but it will perform better than assuming an optimal human, and as we notice failure modes we can improve the model. In the language of this post by Jacob Steinhardt (original  here), we are using a mis-specified human model. The post talks about why model mis-specification is worse than it may seem at first glance. This post is fairly technical and may not be accessible if you don’t have a background in machine learning. If so, you can skip this post and still understand the rest of the posts in the sequence. However, if you want to do ML-related safety research, I strongly recommend putting in the effort to understand the problems that can arise with mis-specification. -------------------------------------------------------------------------------- Machine learning is very good at optimizing predictions to match an observed signal — for instance, given a dataset of input images and labels of the images (e.g. dog, cat, etc.), machine learning is very good at correctly predicting the label of a new image. However, performance can quickly break down as soon as we care about criteria other than predicting observables. There are several cases where we might care about such criteria:  * In scientific investigations, we often care less about predicting a specific    observable phenomenon, and more about what that phenomenon implies about an    underlying scientific theory.  * In economic analysis, we are most interested in what policies will lead to    desirable outcomes. This requires predicting what would counterfactually    happen if we were to enact the policy, which we (usually) don’t have any data    about.  * In ma",https://www.alignmentforum.org/posts/gnvrixhDfG7S2TpNL/latent-variables-and-model-mis-specification,2018,blogPost,"Steinhardt, Jacob",AI Alignment Forum,TAI safety research
+Economics of the singularity,,http://ieeexplore.ieee.org/document/4531461/,2008,journalArticle,"Hanson, Robin",IEEE Spectrum,TAI safety research
+Penalizing side effects using stepwise relative reachability,"How can we design safe reinforcement learning agents that avoid unnecessary disruptions to their environment? We show that current approaches to penalizing side effects can introduce bad incentives, e.g. to prevent any irreversible changes in the environment, including the actions of other agents. To isolate the source of such undesirable incentives, we break down side effects penalties into two components: a baseline state and a measure of deviation from this baseline state. We argue that some of these incentives arise from the choice of baseline, and others arise from the choice of deviation measure. We introduce a new variant of the stepwise inaction baseline and a new deviation measure based on relative reachability of states. The combination of these design choices avoids the given undesirable incentives, while simpler baselines and the unreachability measure fail. We demonstrate this empirically by comparing different combinations of baseline and deviation measure choices on a set of gridworld experiments designed to illustrate possible bad incentives.",http://arxiv.org/abs/1806.01186,2019,conferencePaper,"Krakovna, Victoria; Orseau, Laurent; Kumar, Ramana; Martic, Miljan; Legg, Shane",Proceedings of the Workshop on Artificial Intelligence Safety 2019,TAI safety research
+“Explaining” machine learning reveals policy challenges,,https://www.sciencemag.org/lookup/doi/10.1126/science.aba9647,2020,journalArticle,"Coyle, Diane; Weller, Adrian",Science,TAI safety research
+How unlikely is a doomsday catastrophe?,"Numerous Earth-destroying doomsday scenarios have recently been analyzed, including breakdown of a metastable vacuum state and planetary destruction triggered by a ""strangelet'' or microscopic black hole. We point out that many previous bounds on their frequency give a false sense of security: one cannot infer that such events are rare from the the fact that Earth has survived for so long, because observers are by definition in places lucky enough to have avoided destruction. We derive a new upper bound of one per 10^9 years (99.9% c.l.) on the exogenous terminal catastrophe rate that is free of such selection bias, using planetary age distributions and the relatively late formation time of Earth.",https://arxiv.org/abs/astro-ph/0512204v2,2005,manuscript,"Tegmark, Max; Bostrom, Nick",,TAI safety research
+A new model and dataset for long-range memory,"This blog introduces a new long-range memory model, the Compressive Transformer, alongside a new benchmark for book-level language modelling, PG19. We provide the conceptual tools needed to understand this new research in the context of recent developments in memory models and language modelling.",deepmind.com/blog/article/A_new_model_and_dataset_for_long-range_memory,2020,blogPost,"Rae, Jack; Lillicrap, Timothy",Deepmind,not TAI safety research
+Safe Imitation Learning via Fast Bayesian Reward Inference from Preferences,"Bayesian reward learning from demonstrations enables rigorous safety and uncertainty analysis when performing imitation learning. However, Bayesian reward learning methods are typically computationally intractable for complex control problems. We propose Bayesian Reward Extrapolation (Bayesian REX), a highly efficient Bayesian reward learning algorithm that scales to high-dimensional imitation learning problems by pre-training a low-dimensional feature encoding via self-supervised tasks and then leveraging preferences over demonstrations to perform fast Bayesian inference. Bayesian REX can learn to play Atari games from demonstrations, without access to the game score and can generate 100,000 samples from the posterior over reward functions in only 5 minutes on a personal laptop. Bayesian REX also results in imitation learning performance that is competitive with or better than stateof-the-art methods that only learn point estimates of the reward function. Finally, Bayesian REX enables efficient high-confidence policy evaluation without having access to samples of the reward function. These high-confidence performance bounds can be used to rank the performance and risk of a variety of evaluation policies and provide a way to detect reward hacking behaviors.",http://arxiv.org/abs/2002.09089,2020,conferencePaper,"Brown, Daniel S.; Coleman, Russell; Srinivasan, Ravi; Niekum, Scott","arXiv:2002.09089 [cs, stat]",TAI safety research
+Specification gaming: the flip side of AI ingenuity,"Specification gaming is a behaviour that satisfies the literal specification of an objective without achieving the intended outcome. We have all had experiences with specification gaming, even if not by this name. Readers may have heard the myth of King Midas and the golden touch, in which the king asks that anything he touches be turned to gold - but soon finds that even food and drink turn to metal in his hands. In the real world, when rewarded for doing well on a homework assignment, a student might copy another student to get the right answers, rather than learning the material - and thus exploit a loophole in the task specification.",deepmind.com/blog/article/Specification-gaming-the-flip-side-of-AI-ingenuity,2020,blogPost,"Krakovna, Victoria; Uesato, Jonathan; Mikulik, Vladimir; Rahtz, Matthew; Everitt, Tom; Kumar, Ramana; Kenton, Zachary; Leike, Jan; Legg, Shane",Deepmind,TAI safety research
+Vingean Reflection: Reliable Reasoning for Self-Improving Agents,"Today, human-level machine intelligence is in the domain of futurism, but there is every reason to expect that it will be developed eventually. Once artificial agents become able to improve themselves further, they may far surpass human intelligence, making it vitally important to ensure that the result of an “intelligence explosion” is aligned with human interests. In this paper, we discuss one aspect of this challenge: ensuring that the initial agent’s reasoning about its future versions is reliable, even if these future versions are far more intelligent than the current reasoner. We refer to reasoning of this sort as Vingean reflection.",https://intelligence.org/files/VingeanReflection.pdf,2015,report,"Fallenstein, Benja; Soares, Nate",,TAI safety research
+Directed Policy Gradient for Safe Reinforcement Learning with Human Advice,"Many currently deployed Reinforcement Learning agents work in an environment shared with humans, be them co-workers, users or clients. It is desirable that these agents adjust to people's preferences, learn faster thanks to their help, and act safely around them. We argue that most current approaches that learn from human feedback are unsafe: rewarding or punishing the agent a-posteriori cannot immediately prevent it from wrong-doing. In this paper, we extend Policy Gradient to make it robust to external directives, that would otherwise break the fundamentally on-policy nature of Policy Gradient. Our technique, Directed Policy Gradient (DPG), allows a teacher or backup policy to override the agent before it acts undesirably, while allowing the agent to leverage human advice or directives to learn faster. Our experiments demonstrate that DPG makes the agent learn much faster than reward-based approaches, while requiring an order of magnitude less advice.",http://arxiv.org/abs/1808.04096,2018,manuscript,"Plisnier, Hélène; Steckelmacher, Denis; Brys, Tim; Roijers, Diederik M.; Nowé, Ann",,TAI safety research
+Cognitive prostheses for goal achievement,"Procrastination takes a considerable toll on people’s lives, the economy and society at large. Procrastination is often a consequence of people’s propensity to prioritize their immediate experiences over the long-term consequences of their actions. This suggests that aligning immediate rewards with long-term values could be a promising way to help people make more future-minded decisions and overcome procrastination. Here we develop an approach to decision support that leverages artificial intelligence and game elements to restructure challenging sequential decision problems in such a way that it becomes easier for people to take the right course of action. A series of four increasingly realistic experiments suggests that this approach can enable people to make better decisions faster, procrastinate less, complete their work on time and waste less time on unimportant tasks. These findings suggest that our method is a promising step towards developing cognitive prostheses that help people achieve their goals.",https://www.nature.com/articles/s41562-019-0672-9,2019,journalArticle,"Lieder, Falk; Chen, Owen X.; Krueger, Paul M.; Griffiths, Thomas L.",Nature Human Behaviour,not TAI safety research
+Forecasting Transformative AI: An Expert Survey,"Transformative AI technologies have the potential to reshape critical aspects of society in the near future. However, in order to properly prepare policy initiatives for the arrival of such technologies accurate forecasts and timelines are necessary. A survey was administered to attendees of three AI conferences during the summer of 2018 (ICML, IJCAI and the HLAI conference). The survey included questions for estimating AI capabilities over the next decade, questions for forecasting five scenarios of transformative AI and questions concerning the impact of computational resources in AI research. Respondents indicated a median of 21.5% of human tasks (i.e., all tasks that humans are currently paid to do) can be feasibly automated now, and that this figure would rise to 40% in 5 years and 60% in 10 years. Median forecasts indicated a 50% probability of AI systems being capable of automating 90% of current human tasks in 25 years and 99% of current human tasks in 50 years. The conference of attendance was found to have a statistically significant impact on all forecasts, with attendees of HLAI providing more optimistic timelines with less uncertainty. These findings suggest that AI experts expect major advances in AI technology to continue over the next decade to a degree that will likely have profound transformative impacts on society.",http://arxiv.org/abs/1901.08579,2019,manuscript,"Gruetzemacher, Ross; Paradice, David; Lee, Kang Bok",,TAI safety research
+Guide Me: Interacting with Deep Networks,"Interaction and collaboration between humans and intelligent machines has become increasingly important as machine learning methods move into real-world applications that involve end users. While much prior work lies at the intersection of natural language and vision, such as image captioning or image generation from text descriptions, less focus has been placed on the use of language to guide or improve the performance of a learned visual processing algorithm. In this paper, we explore methods to flexibly guide a trained convolutional neural network through user input to improve its performance during inference. We do so by inserting a layer that acts as a spatio-semantic guide into the network. This guide is trained to modify the network's activations, either directly via an energy minimization scheme or indirectly through a recurrent model that translates human language queries to interaction weights. Learning the verbal interaction is fully automatic and does not require manual text annotations. We evaluate the method on two datasets, showing that guiding a pre-trained network can improve performance, and provide extensive insights into the interaction between the guide and the CNN.",http://arxiv.org/abs/1803.11544,2018,conferencePaper,"Rupprecht, Christian; Laina, Iro; Navab, Nassir; Hager, Gregory D.; Tombari, Federico",Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR),not TAI safety research
+Thread: Circuits,What can we learn if we invest heavily in reverse engineering a single neural network?,https://distill.pub/2020/circuits,2020,journalArticle,"Cammarata, Nick; Carter, Shan; Goh, Gabriel; Olah, Chris; Petrov, Michael; Schubert, Ludwig",Distill,not TAI safety research
+Visualizing Representations: Deep Learning and Human Beings - colah's blog,,http://colah.github.io/posts/2015-01-Visualizing-Representations/,2015,blogPost,"Olah, Chris",Colah's blog,not TAI safety research
+One Decade of Universal Artificial Intelligence,"The first decade of this century has seen the nascency of the first mathematical theory of general artificial intelligence. This theory of Universal Artificial Intelligence (UAI) has made significant contributions to many theoretical, philosophical, and practical AI questions. In a series of papers culminating in book (Hutter, 2005), an exciting sound and complete mathematical model for a super intelligent agent (AIXI) has been developed and rigorously analyzed. While nowadays most AI researchers avoid discussing intelligence, the award-winning PhD thesis (Legg, 2008) provided the philosophical embedding and investigated the UAI-based universal measure of rational intelligence, which is formal, objective and non-anthropocentric. Recently, effective approximations of AIXI have been derived and experimentally investigated in JAIR paper (Veness et al. 2011). This practical breakthrough has resulted in some impressive applications, finally muting earlier critique that UAI is only a theory. For the first time, without providing any domain knowledge, the same agent is able to self-adapt to a diverse range of interactive environments. For instance, AIXI is able to learn from scratch to play TicTacToe, Pacman, Kuhn Poker, and other games by trial and error, without even providing the rules of the games. These achievements give new hope that the grand goal of Artificial General Intelligence is not elusive. This article provides an informal overview of UAI in context. It attempts to gently introduce a very theoretical, formal, and mathematical subject, and discusses philosophical and technical ingredients, traits of intelligence, some social questions, and the past and future of UAI.",http://arxiv.org/abs/1202.6153,2012,journalArticle,"Hutter, Marcus",Theoretical Foundations of Artificial General Intelligence,TAI safety research
+Should Artificial Intelligence Governance be Centralised?: Design Lessons from History,,https://dl.acm.org/doi/10.1145/3375627.3375857,2020,conferencePaper,"Cihon, Peter; Maas, Matthijs M.; Kemp, Luke","Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society",TAI safety research
+Feature Expansive Reward Learning: Rethinking Human Input,"In collaborative human-robot scenarios, when a person is not satisfied with how a robot performs a task, they can intervene to correct it. Reward learning methods enable the robot to adapt its reward function online based on such human input. However, due to the real-time nature of the input, this online adaptation requires low sample complexity algorithms which rely on simple functions of handcrafted features. In practice, pre-specifying an exhaustive set of features the person might care about is impossible; what should the robot do when the human correction cannot be explained by the features it already has access to? Recent progress in deep Inverse Reinforcement Learning (IRL) suggests that the robot could fall back on demonstrations: ask the human for demonstrations of the task, and recover a reward defined over not just the known features, but also the raw state space. Our insight is that rather than implicitly learning about the missing feature(s) from task demonstrations, the robot should instead ask for data that explicitly teaches it about what it is missing. We introduce a new type of human input, in which the person guides the robot from areas of the state space where the feature she is teaching is highly expressed to states where it is not. We propose an algorithm for learning the feature from the raw state space and integrating it into the reward function. By focusing the human input on the missing feature, our method decreases sample complexity and improves generalization of the learned reward over the above deep IRL baseline. We show this in experiments with a 7DOF robot manipulator. Finally, we discuss our method’s potential implications for deep reward learning more broadly: taking a divide-and-conquer approach that focuses on important features separately before learning from demonstrations can improve generalization in tasks where such features are easy for the human to teach.",http://arxiv.org/abs/2006.13208,2020,manuscript,"Bobu, Andreea; Wiggert, Marius; Tomlin, Claire; Dragan, Anca D.",,not TAI safety research
+Emergent Complexity via Multi-Agent Competition,"Reinforcement learning algorithms can train agents that solve problems in complex, interesting environments. Normally, the complexity of the trained agent is closely related to the complexity of the environment. This suggests that a highly capable agent requires a complex environment for training. In this paper, we point out that a competitive multi-agent environment trained with self-play can produce behaviors that are far more complex than the environment itself. We also point out that such environments come with a natural curriculum, because for any skill level, an environment full of agents of this level will have the right level of difficulty. This work introduces several competitive multi-agent environments where agents compete in a 3D world with simulated physics. The trained agents learn a wide variety of complex and interesting skills, even though the environment themselves are relatively simple. The skills include behaviors such as running, blocking, ducking, tackling, fooling opponents, kicking, and defending using both arms and legs. A highlight of the learned behaviors can be found here: https://goo.gl/eR7fbX",http://arxiv.org/abs/1710.03748,2018,conferencePaper,"Bansal, Trapit; Pachocki, Jakub; Sidor, Szymon; Sutskever, Ilya; Mordatch, Igor",arXiv:1710.03748 [cs],not TAI safety research
+Learning Agile Robotic Locomotion Skills by Imitating Animals,"Reproducing the diverse and agile locomotion skills of animals has been a longstanding challenge in robotics. While manually-designed controllers have been able to emulate many complex behaviors, building such controllers involves a time-consuming and difficult development process, often requiring substantial expertise of the nuances of each skill. Reinforcement learning provides an appealing alternative for automating the manual effort involved in the development of controllers. However, designing learning objectives that elicit the desired behaviors from an agent can also require a great deal of skill-specific expertise. In this work, we present an imitation learning system that enables legged robots to learn agile locomotion skills by imitating real-world animals. We show that by leveraging reference motion data, a single learning-based approach is able to automatically synthesize controllers for a diverse repertoire behaviors for legged robots. By incorporating sample efficient domain adaptation techniques into the training process, our system is able to learn adaptive policies in simulation that can then be quickly adapted for real-world deployment. To demonstrate the effectiveness of our system, we train an 18-DoF quadruped robot to perform a variety of agile behaviors ranging from different locomotion gaits to dynamic hops and turns.",http://arxiv.org/abs/2004.00784,2020,conferencePaper,"Peng, Xue Bin; Coumans, Erwin; Zhang, Tingnan; Lee, Tsang-Wei; Tan, Jie; Levine, Sergey",arXiv:2004.00784 [cs],not TAI safety research
+Antitrust-Compliant AI Industry Self-Regulation,"The touchstone of antitrust compliance is competition. To be legally permissible, any industrial restraint on trade must have sufficient countervailing procompetitive justifications. Usually, anticompetitive horizontal agreements like boycotts (including a refusal to produce certain products) are per se illegal.",https://cullenokeefe.com/blog/antitrust-compliant-ai-industry-self-regulation,2020,manuscript,"O’Keefe, Cullen",,TAI safety research
+Machine Learning Explainability for External Stakeholders,"As machine learning is increasingly deployed in high-stakes contexts affecting people's livelihoods, there have been growing calls to open the black box and to make machine learning algorithms more explainable. Providing useful explanations requires careful consideration of the needs of stakeholders, including end-users, regulators, and domain experts. Despite this need, little work has been done to facilitate inter-stakeholder conversation around explainable machine learning. To help address this gap, we conducted a closed-door, day-long workshop between academics, industry experts, legal scholars, and policymakers to develop a shared language around explainability and to understand the current shortcomings of and potential solutions for deploying explainable machine learning in service of transparency goals. We also asked participants to share case studies in deploying explainable machine learning at scale. In this paper, we provide a short summary of various case studies of explainable machine learning, lessons from those studies, and discuss open challenges.",https://arxiv.org/abs/2007.05408v1,2020,conferencePaper,"Bhatt, Umang; Andrus, McKane; Weller, Adrian; Xiang, Alice",,TAI safety research
+Avoiding Wireheading with Value Reinforcement Learning,"How can we design good goals for arbitrarily intelligent agents? Reinforcement learning (RL) is a natural approach. Unfortunately, RL does not work well for generally intelligent agents, as RL agents are incentivised to shortcut the reward sensor for maximum reward -- the so-called wireheading problem. In this paper we suggest an alternative to RL called value reinforcement learning (VRL). In VRL, agents use the reward signal to learn a utility function. The VRL setup allows us to remove the incentive to wirehead by placing a constraint on the agent's actions. The constraint is defined in terms of the agent's belief distributions, and does not require an explicit specification of which actions constitute wireheading.",http://arxiv.org/abs/1605.03143,2016,conferencePaper,"Everitt, Tom; Hutter, Marcus",AGI 2016: Artificial General Intelligence,TAI safety research
+Principles for the Application of Human Intelligence,"Before humans become the standard way in which we make decisions, we need to consider the risks and ensure implementation of human decision-making systems does not cause widespread harm.",https://behavioralscientist.org/principles-for-the-application-of-human-intelligence/,2019,blogPost,"Collins, Jason",Behavioral Scientist,not TAI safety research
+Backup utility functions as a fail-safe AI technique,"Many experts believe that AIs will, within the not-too-distant future, become powerful enough for their decisions to have tremendous impact. Unfortunately, setting up AI goal systems in a way that results in benevolent behavior is expected to be difficult, and we cannot be certain to get it completely right on the first attempt. We should therefore account for the possibility that the goal systems fail to implement our values the intended way. In this paper, we propose the idea of backup utility functions: Secondary utility functions that are used in case the primary ones “fail”. We also describe how this approach can be generalized to the use of multi-layered utility functions, some of which can fail without affecting the final outcome as badly as without the backup mechanism.",https://longtermrisk.org/files/backup-utility-functions.pdf,2016,manuscript,"Oesterheld, Caspar",,TAI safety research
+Predicting human decisions with behavioral theories and machine learning,"Behavioral decision theories aim to explain human behavior. Can they help predict it? An open tournament for prediction of human choices in fundamental economic decision tasks is presented. The results suggest that integration of certain behavioral theories as features in machine learning systems provides the best predictions. Surprisingly, the most useful theories for prediction build on basic properties of human and animal learning and are very different from mainstream decision theories that focus on deviations from rational choice. Moreover, we find that theoretical features should be based not only on qualitative behavioral insights (e.g. loss aversion), but also on quantitative behavioral foresights generated by functional descriptive models (e.g. Prospect Theory). Our analysis prescribes a recipe for derivation of explainable, useful predictions of human decisions.",http://arxiv.org/abs/1904.06866,2019,manuscript,"Plonsky, Ori; Apel, Reut; Ert, Eyal; Tennenholtz, Moshe; Bourgin, David; Peterson, Joshua C.; Reichman, Daniel; Griffiths, Thomas L.; Russell, Stuart J.; Carter, Evan C.; Cavanagh, James F.; Erev, Ido",,TAI safety research
+"Exchange-Traded Funds, Market Structure, and the Flash Crash",,https://www.tandfonline.com/doi/full/10.2469/faj.v68.n4.6,2012,journalArticle,"Madhavan, Ananth",Financial Analysts Journal,not TAI safety research
+A general model of safety-oriented AI development,"This may be trivial or obvious for a lot of people, but it doesn't seem like anyone has bothered to write it down (or I haven't looked hard enough). It started out as a generalization of Paul Christiano's IDA, but also covers things like safe recursive self-improvement. Start with a team of one or more humans (researchers, programmers, trainers, and/or overseers), with access to zero or more AIs (initially as assistants). The human/AI team in each round develops a new AI and adds it to the team, and repeats this until maturity in AI technology is achieved. Safety/alignment is ensured by having some set of safety/alignment properties on the team that is inductively maintained by the development process. The reason I started thinking in this direction is that Paul's approach seemed very hard to knock down, because any time a flaw or difficulty is pointed out or someone expresses skepticism on some technique that it uses or the overall safety invariant, there's always a list of other techniques or invariants that could be substituted in for that part (sometimes in my own brain as I tried to criticize some part of it). Eventually I realized this shouldn't be surprising because IDA is an instance of this more general model of safety-oriented AI development, so there are bound to be many points near it in the space of possible safety-oriented AI development practices. (Again, this may already be obvious to others including Paul, and in their minds IDA is perhaps already a cluster of possible development practices consisting of the most promising safety techniques and invariants, rather than a single point.) If this model turns out not to have been written down before, perhaps it should be assigned a name, like Iterated Safety-Invariant AI-Assisted AI Development, or something pithier?",https://www.alignmentforum.org/posts/idb5Ppp9zghcichJ5/a-general-model-of-safety-oriented-ai-development,2018,blogPost,Wei Dai,AI Alignment Forum,TAI safety research
+The Role and Limits of Principles in AI Ethics: Towards a Focus on Tensions,"The last few years have seen a proliferation of principles for AI ethics. There is substantial overlap between different sets of principles, with widespread agreement that AI should be used for the common good, should not be used to harm people or undermine their rights, and should respect widely held values such as fairness, privacy, and autonomy. While articulating and agreeing on principles is important, it is only a starting point. Drawing on comparisons with the field of bioethics, we highlight some of the limitations of principles: in particular, they are often too broad and high-level to guide ethics in practice. We suggest that an important next step for the field of AI ethics is to focus on exploring the tensions that inevitably arise as we try to implement principles in practice. By explicitly recognising these tensions we can begin to make decisions about how they should be resolved in specific cases, and develop frameworks and guidelines for AI ethics that are rigorous and practically relevant. We discuss some different specific ways that tensions arise in AI ethics, and what processes might be needed to resolve them.",,2019,conferencePaper,"Whittlestone, Jess; Nyrup, Rune; Alexandrova, Anna; Cave, Stephen","AIES '19: Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society",TAI safety research
+Enhancing metacognitive reinforcement learning using reward structures and feedback,"How do we learn to think better, and what can we do to promote such metacognitive learning? Here, we propose that cognitive growth proceeds through metacognitive reinforcement learning. We apply this theory to model how people learn how far to plan ahead and test its predictions about the speed of metacognitive learning in two experiments. In the first experiment, we find that our model can discern a reward structure that promotes metacognitive reinforcement learning from one that hinders it. In the second experiment, we show that our model can be used to design a feedback mechanism that enhances metacognitive reinforcement learning in an environment that hinders learning. Our results suggest that modeling metacognitive learning is a promising step towards promoting cognitive growth.",,2017,conferencePaper,"Krueger, Paul M; Lieder, Falk; Griffiths, Thomas L",39th Annual Meeting of the Cognitive Science Society,not TAI safety research
+Learning agents for uncertain environments (extended abstract),,http://portal.acm.org/citation.cfm?doid=279943.279964,1998,conferencePaper,"Russell, Stuart",Proceedings of the eleventh annual conference on Computational learning theory  - COLT' 98,TAI safety research
+Existential Risk and Growth,"Human activity can create or mitigate risks of catastrophes, such as nuclear war, climate change, pandemics, or artificial intelligence run amok. These could even imperil the survival of human civilization. What is the relationship between economic growth and such existential risks? In a model of directed technical change, with moderate parameters, existential risk follows a Kuznets-style inverted U-shape. This suggests we could be living in a unique “time of perils,” having developed technologies advanced enough to threaten our permanent destruction, but not having grown wealthy enough yet to be willing to spend sufficiently on safety. Accelerating growth during this “time of perils” initially increases risk, but improves the chances of humanity’s survival in the long run. Conversely, even short-term stagnation could substantially curtail the future of humanity.",,2020,report,"Aschenbrenner, Leopold",,not TAI safety research
+Coherence arguments do not imply goal-directed behavior,"One of the most pleasing things about probability and expected utility theory is that there are many coherence arguments that suggest that these are the “correct” ways to reason. If you deviate from what the theory prescribes, then you must be executing a dominated strategy. There must be some other strategy that never does any worse than your strategy, but does strictly better than your strategy with certainty in at least one situation. There’s a good explanation of these arguments here. We shouldn’t expect mere humans to be able to notice any failures of coherence in a superintelligent agent, since if we could notice these failures, so could the agent. So we should expect that powerful agents appear coherent to us. (Note that it is possible that the agent doesn’t fix the failures because it would not be worth it -- in this case, the argument says that we will not be able to notice any exploitable failures.) Taken together, these arguments suggest that we should model an agent much smarter than us as an expected utility (EU) maximizer. And many people agree that EU maximizers are dangerous. So does this mean we’re doomed? I don’t think so: it seems to me that the problems about EU maximizers that we’ve identified are actually about goal-directed behavior or explicit reward maximizers. The coherence theorems say nothing about whether an AI system must look like one of these categories. This suggests that we could try building an AI system that can be modeled as an EU maximizer, yet doesn’t fall into one of these two categories, and so doesn’t have all of the problems that we worry about. Note that there are two different flavors of arguments that the AI systems we build will be goal-directed agents (which are dangerous if the goal is even slightly wrong):  * Simply knowing that an agent is intelligent lets us infer that it is    goal-directed. (ETA: See this comment for more details on this argument.)  * Humans are particularly likely to build goal-directed agen",https://www.alignmentforum.org/posts/NxF5G6CJiof6cemTw/coherence-arguments-do-not-imply-goal-directed-behavior,2018,blogPost,"Shah, Rohin",AI Alignment Forum,TAI safety research
+Two Alternatives to Logical Counterfactuals,"The following is a critique of the idea of logical counterfactuals. The idea of logical counterfactuals has appeared in previous agent foundations research (especially at MIRI): here, here. “…",https://unstableontology.com/2020/04/01/alternatives-to-logical-counterfactuals/,2020,blogPost,"Taylor, Jessica",Unstable Ontology,TAI safety research
+The race for an artificial general intelligence: implications for public policy,"An arms race for an artificial general intelligence (AGI) would be detrimental for and even pose an existential threat to humanity if it results in an unfriendly AGI. In this paper, an all-pay contest model is developed to derive implications for public policy to avoid such an outcome. It is established that, in a winner-takes-all race, where players must invest in R&D, only the most competitive teams will participate. Thus, given the difficulty of AGI, the number of competing teams is unlikely ever to be very large. It is also established that the intention of teams competing in an AGI race, as well as the possibility of an intermediate outcome (prize), is important. The possibility of an intermediate prize will raise the probability of finding the dominant AGI application and, hence, will make public control more urgent. It is recommended that the danger of an unfriendly AGI can be reduced by taxing AI and using public procurement. This would reduce the pay-off of contestants, raise the amount of R&D needed to compete, and coordinate and incentivize co-operation. This will help to alleviate the control and political problems in AI. Future research is needed to elaborate the design of systems of public procurement of AI innovation and for appropriately adjusting the legal frameworks underpinning high-tech innovation, in particular dealing with patenting by AI.",https://doi.org/10.1007/s00146-019-00887-x,2019,journalArticle,"Naudé, Wim; Dimitri, Nicola",AI & Society,TAI safety research
+Neuroevolution of Self-Interpretable Agents,"Inattentional blindness is the psychological phenomenon that causes one to miss things in plain sight. It is a consequence of the selective attention in perception that lets us remain focused on important parts of our world without distraction from irrelevant details. Motivated by selective attention, we study the properties of artificial agents that perceive the world through the lens of a self-attention bottleneck. By constraining access to only a small fraction of the visual input, we show that their policies are directly interpretable in pixel space. We find neuroevolution ideal for training self-attention architectures for vision-based reinforcement learning (RL) tasks, allowing us to incorporate modules that can include discrete, non-differentiable operations which are useful for our agent. We argue that self-attention has similar properties as indirect encoding, in the sense that large implicit weight matrices are generated from a small number of key-query parameters, thus enabling our agent to solve challenging vision based tasks with at least 1000x fewer parameters than existing methods. Since our agent attends to only task critical visual hints, they are able to generalize to environments where task irrelevant elements are modified while conventional methods fail. Videos of our results and source code available at https://attentionagent.github.io/",http://arxiv.org/abs/2003.08165,2020,conferencePaper,"Tang, Yujin; Nguyen, Duong; Ha, David",Proceedings of the 2020 Genetic and Evolutionary Computation Conference,not TAI safety research
+Brainjacking in deep brain stimulation and autonomy,,,2018,journalArticle,"Pugh, Jonathan; Pycroft, Laurie; Sandberg, Anders; Aziz, Tipu; Savulescu, Julian",Ethics and information technology,not TAI safety research
+AI development incentive gradients are not uniformly terrible,"Much of the work for this post was done together with Nuño Sempere Perhaps you think that your values will be best served if the AGI you (or your team, company or nation) are developing is deployed first. Would you decide that it's worth cutting a few corners, reducing your safety budget, and pushing ahead to try and get your AI out the door first? It seems plausible, and worrying, that you might. And if your competitors reason symmetrically, we would get a ""safety race to the bottom"". On the other hand, perhaps you think your values will be better served if your enemy wins than if either of you accidentally produces an unfriendly AI. Would you decide the safety costs to improving your chances aren't worth it? In a simple two player model, you should only shift funds from safety to capabilities if (the relative₁ decrease in chance of friendliness) / (the relative₁ increase in the chance of winning) < (expected relative₂ loss of value if your enemy wins rather than you). Here, the relative₁ increases and decreases are relative to the current values. The relative₂ loss of value is relative to the expected value if you win. The plan of this post is as follows: 1. Consider a very simple model that leads to a safety race. Identify unrealistic assumptions which are driving its results. 2. Remove some of the unrealistic assumptions and generate a different model. Derive the inequality expressed above. 3. Look at some specific example cases, and see how they affect safety considerations. A PARTLY DISCONTINUOUS MODEL Let's consider a model with two players with the same amount of resources. Each player's choice is what fraction of their resources to devote to safety, rather than capabilities. Whichever player contributes more to capabilities wins the race. If you win the race, you either get a good outcome or a bad outcome. Your chance of getting a good outcome increases continuously with the amount you spent on safety. If the other player wins, you get a bad outcome.",https://www.lesswrong.com/posts/bkG4qj9BFEkNva3EX/ai-development-incentive-gradients-are-not-uniformly,2018,blogPost,rk,LessWrong,TAI safety research
+What is ambitious value learning?,"I think of ambitious value learning as a proposed solution to the specification problem, which I define as the problem of defining the behavior that we would want to see from our AI system. I italicize “defining” to emphasize that this is  not the problem of actually computing behavior that we want to see -- that’s the full AI safety problem. Here we are allowed to use hopelessly impractical schemes, as long as the resulting definition would allow us to in theory compute the behavior that an AI system would take, perhaps with assumptions like infinite computing power or arbitrarily many queries to a human. (Although we do prefer specifications that seem like they could admit an efficient implementation.) In terms of DeepMind’s classification, we are looking for a design specification that exactly matches the ideal specification. HCH and  indirect normativity are examples of attempts at such specifications. We will consider a model in which our AI system is maximizing the expected utility of some explicitly represented utility function that can depend on history. (It does not matter materially whether we consider utility functions or reward functions, as long as they can depend on history.) The utility function may be learned from data, or designed by hand, but it must be an explicit part of the AI that is then maximized. I will not justify this model for now, but simply assume it by fiat and see where it takes us. I’ll note briefly that this model is often justified by the  VNM utility theorem and AIXI, and as the natural idealization of reinforcement learning, which aims to maximize the expected sum of rewards, although typically rewards in RL depend only on states. A lot of conceptual arguments, as well as experiences with specification gaming, suggest that we are unlikely to be able to simply think hard and write down a good specification, since even small errors in specifications can lead to bad results. However, machine learning is particularly good at narro",https://www.alignmentforum.org/posts/5eX8ko7GCxwR5N9mN/what-is-ambitious-value-learning,2018,blogPost,"Shah, Rohin",AI Alignment Forum,TAI safety research

data/twitter_complaints/task.json

@@ -0,0 +1 @@
+{"name": "twitter_complaints", "description": "", "data_columns": ["Tweet text"], "label_columns": {"Label": ["complaint", "no complaint"]}}

data/twitter_complaints/train.csv

@@ -0,0 +1,51 @@
+Tweet text,Label
+@HMRCcustomers No this is my first job,no complaint
+"@KristaMariePark Thank you for your interest! If you decide to cancel, you can call Customer Care at 1-800-NYTIMES.",no complaint
+If I can't get my 3rd pair of @beatsbydre powerbeats to work today I'm doneski man. This is a slap in my balls. Your next @Bose @BoseService,complaint
+@EE On Rosneath Arial having good upload and download speeds but terrible latency 200ms. Why is this.,complaint
+"Couples wallpaper, so cute. :) #BrothersAtHome",no complaint
+"@mckelldogs This might just be me, but-- eyedrops? Artificial tears are so useful when you're sleep-deprived and sp… https://t.co/WRtNsokblG",no complaint
+@Yelp can we get the exact calculations for a business rating (for example if its 4 stars but actually 4.2) or do we use a 3rd party site?,no complaint
+@nationalgridus I have no water and the bill is current and paid. Can you do something about this?,complaint
+"Never shopping at @MACcosmetics again. Every time I go in there, their employees are super rude/condescending. I'll take my $$ to @Sephora",complaint
+@JenniferTilly Merry Christmas to as well. You get more stunning every year ��,no complaint
+@NortonSupport Thanks much.,no complaint
+@VerizonSupport all of a sudden I can't connect to my primary wireless network but guest one works,no complaint
+Aaaahhhhh!!!! My @Razer @PlayOverwatch d.va meka headset came in!!! I didn't even know it had shipped!!! So excited… https://t.co/4gXy9xED8d,no complaint
+@Lin_Manuel @jmessinaphoto @VAMNit Omg a little squish!!!!! Enjoy and congrats!!!! I miss mine being so young! ������,no complaint
+@IanJamesPoulter What's your secret to poaching eggs? Mine NEVER look that good.,no complaint
+@AWSSupport When will be able Kinesis Firehose compatible with Elasticsearch 6.0? Thank you!,no complaint
+@NCIS_CBS https://t.co/eeVL9Eu3bE,no complaint
+@msetchell Via the settings? That’s how I do it on master T’s,no complaint
+"Today at work there was a low flying duck heading toward a crowd of people, and I yelled ""watch out! and I'm very disappointed with myself.",no complaint
+@NortonSupport @NortonOnline What the hell is a dm 5-10 days to get money back bank account now overdrawn thanks guys,complaint
+@united not happy with this delay from Newark to Manchester tonight :( only 30 mins free Wi-fi sucks ...,complaint
+@ZARA_Care I've been waiting on a reply to my tweets and DMs for days now?,complaint
+"New Listing! Large 2 Family Home for Sale in #Passaic Park, #NJ #realestate #homesforsale Great Location!… https://t.co/IV4OrLXkMk",no complaint
+@SouthwestAir I love you but when sending me flight changes please don't use military time #ignoranceisbliss,complaint
+@JetBlue Completely understand but would prefer being on time to filling out forms....,no complaint
+@nvidiacc I own two gtx 460 in sli. I want to try windows 8 dev preview. Which driver should I use. Can I use the windows 7 one.,no complaint
+Just posted a photo https://t.co/RShFwCjPHu,no complaint
+Love crescent rolls? Try adding pesto @PerdueChicken to them and you’re going to love it! #Promotion #PerdueCrew -… https://t.co/KBHOfqCukH,no complaint
+@TopmanAskUs please just give me my money back.,complaint
+I just gave 5 stars to Tracee at @neimanmarcus for the great service I received!,no complaint
+@FitbitSupport when are you launching new clock faces for Indian market,no complaint
+@HPSupport my printer will not allow me to choose color instead it only prints monochrome #hppsdr #ijkhelp,complaint
+@DIRECTV can I get a monthly charge double refund when it sprinkles outside and we lose reception? #IamEmbarrasedForYou,complaint
+"@AlfaRomeoCares Hi thanks for replying, could be my internet but link doesn't seem to be working",complaint
+Looks tasty! Going to share with everyone I know #FebrezeONE #sponsored https://t.co/4AQI53npei,no complaint
+@OnePlus_IN can OnePlus 5T do front camera portrait?,no complaint
+@sho_help @showtime your arrive is terrible streaming is stop and start every couple mins. Get it together it's xmas,complaint
+@KandraKPTV I just witnessed a huge building fire in Santa Monica California,no complaint
+@fernrocks most definitely the latter for me,no complaint
+@greateranglia Could I ask why the Area in front of BIC Station was not gritted withh all the snow.,complaint
+I'm earning points with #CricketRewards https://t.co/GfpGhqqnhE,no complaint
+@Schrapnel @comcast RIP me,no complaint
+"The wait is finally over, just joined @SquareUK, hope to get started real soon!",no complaint
+@WholeFoods what's the best way to give feedback on a particular store to the regional/national office?,no complaint
+@DanielNewman I honestly would believe anything. People are...too much sometimes.,no complaint
+@asblough Yep! It should send you a notification with your driver’s name and what time they’ll be showing up!,no complaint
+@Wavy2Timez for real,no complaint
+@KenyaPower_Care  no power in south b area... is it scheduled.,complaint
+Honda won't do anything about water leaking in brand new car. Frustrated! @HondaCustSvc @AmericanHonda,complaint
+"@CBSNews @Dodge @ChryslerCares My driver side air bag has been recalled and replaced, but what about the passenger side?",complaint

raft.py

@@ -47,20 +47,11 @@ _LICENSE = ""
 # TODO: Add link to the official dataset URLs here
 # The HuggingFace dataset library don't host the datasets but only point to the original files
 # This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
-_URLs = {
-    'TAISafety': {
-        'train': "./data/TAISafety/train.csv",
-        'test': "./data/TAISafety/test.csv"
-    },
-    'AIInitiatives': {
-        'train': "./data/AIInitiatives/train.csv",
-        'test': "./data/AIInitiatives/test.csv"
-    },
-    'MedicalDomain': {
-        'train': "./data/MedicalDomain/train.csv",
-        'test': "./data/MedicalDomain/test.csv"
-    }
-}  # TODO: Generate these automatically.
+# This gets all folders within the directory named `data`
+DATA_DIRS = next(os.walk('data'))[1]
+
+_URLs = {s: {'train': f".data/{s}/train.csv",
+             'test': f".data/{s}/test_unlabeled.csv"} for s in DATA_DIRS}
 
 
 class Raft(datasets.GeneratorBasedBuilder):
@@ -79,58 +70,34 @@ class Raft(datasets.GeneratorBasedBuilder):
     # You will be able to load one or the other configurations in the following list with
     # data = datasets.load_dataset('my_dataset', 'first_domain')
     # data = datasets.load_dataset('my_dataset', 'second_domain')
-    BUILDER_CONFIGS = [
-        datasets.BuilderConfig(name="TAISafety-binary", version=VERSION,
-                               description="Decide whether the papers focus on AI safety methods."),
-        datasets.BuilderConfig(name="TAISafety-multiclass", version=VERSION,
-                               description="If a paper has AI safety methods, determine if it is meta"
-                                           "safety or technical safety."),
-        datasets.BuilderConfig(name="AIInitiatives-multilabel", version=VERSION,
-                               description="For each initiative, decide which (if any) of Ethics, "
-                                           "Governance, and Social Good apply to the initiative's AI goals"),
-        datasets.BuilderConfig(name="MedicalDomain-multiclass", version=VERSION,
-                               description="Which medical subdomain does a given physician's clinical notes "
-                                           "belong to?"),
-    ]
-
-    DEFAULT_CONFIG_NAME = "TAISafety-binary"  # It's not mandatory to have a default configuration. Just use one if it make sense.
+
+    # TODO: Load task jsons
+
+    tasks = []
+    for sd in DATA_DIRS:
+        with open(os.path.join('data', sd, 'task.json')) as f:
+            task_data = json.loads(f)
+        tasks.append(task_data)
+
+    BUILDER_CONFIGS = [datasets.BuilderConfig(name=td['name'], version=VERSION,
+                       description=td['description']) for td in tasks]
+
+    DEFAULT_CONFIG_NAME = "tai_safety_research"  # It's not mandatory to have a default configuration. Just use one if it make sense.
 
     def _info(self):
+        task = Raft.tasks[self.config.name]
         # TODO: This method specifies the datasets.DatasetInfo object which contains informations and typings for the dataset
-        if self.config.name.startswith("TAISafety"):
-            features = datasets.Features(
-                {
-                    "title": datasets.Value("string"),
-                    "publication": datasets.Value("string"),
-                    "abstract": datasets.Value("string"),
-                    "answer": datasets.Value("string"),
-                }
-            )
-        elif self.config.name.startswith("AIInitiatives"):
-            features = datasets.Features(
-                {
-                    "name": datasets.Value("string"),
-                    "organization": datasets.Value("string"),
-                    "description": datasets.Value("string"),
-                    "sector": datasets.Value("string"),
-                    "scope": datasets.Value("string"),
-                    "audience": datasets.Value("string"),
-                    "stage": datasets.Value("string"),
-                    "date": datasets.Value("string"),
-                    "country": datasets.Value("string"),
-                    "notes": datasets.Value("string"),
-                    "answer_ethics": datasets.Value("string"),
-                    "answer_governance": datasets.Value("string"),
-                    "answer_socialgood": datasets.Value("string"),
-                }
-            )
-        elif self.config.name.startswith("MedicalDomain"):
-            features = datasets.Features(
-                {
-                    "text": datasets.Value("string"),
-                    "answer": datasets.Value("string"),
-                }
-            )
+        data_columns = {col_name: datasets.Value("string") for col_name in
+                        task['data_columns']}
+
+        label_columns = {}
+        for label_name in task['label_columns']:
+            labels = task['label_columns'][label_name]
+            label_columns[label_name] = datasets.ClassLabel(len(labels), labels)
+
+        # Merge dicts
+        features = datasets.Features(**data_columns, **label_columns)
+
         return datasets.DatasetInfo(
             # This is the description that will appear on the datasets page.
             description=_DESCRIPTION,
@@ -174,49 +141,16 @@ class Raft(datasets.GeneratorBasedBuilder):
         # This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
         # The `key` is here for legacy reason (tfds) and is not important in itself.
 
-        dataset, config = self.config.name.split("-")
+        task = Raft.tasks[self.config.name]
+        column_names = task['data_columns'] + list(task['label_columns'])
+        num_labels = len(task['label_columns'])
 
         with open(filepath, encoding="utf-8") as f:
-            csv_reader = csv.reader(
-                f, quotechar='"', delimiter=",", quoting=csv.QUOTE_ALL, skipinitialspace=True
-            )
+            csv_reader = csv.reader(f, quotechar='"', delimiter=",",
+                                    quoting=csv.QUOTE_ALL, skipinitialspace=True)
             for id_, row in enumerate(csv_reader):
                 if id_ == 0:  # First row is column names
                     continue
-                if dataset == "TAISafety":
-                    if split == "train":
-                        title, publication, abstract, category, binary = row
-                        answer = category if config == "multiclass" else binary
-                    elif split == "test":
-                        title, publication, abstract = row
-                        answer = ""
-                    yield id_, {"title": title,
-                                "publication": publication,
-                                "abstract": abstract,
-                                "answer": answer}
-                if dataset == "AIInitiatives":
-                    name, organization, description, sector, scope, audience, \
-                        stage, date, country, notes, answer_ethics, \
-                        answer_governance, answer_socialgood = row
-                    if split == "test":
-                        answer_ethics, answer_governance, answer_socialgood = "", "", ""
-                    yield id_, {"name": name,
-                                "organization": organization,
-                                "description": description,
-                                "sector": sector,
-                                "scope": scope,
-                                "audience": audience,
-                                "stage": stage,
-                                "date": date,
-                                "country": country,
-                                "notes": notes,
-                                "answer_ethics": answer_ethics,
-                                "answer_governance": answer_governance,
-                                "answer_socialgood": answer_socialgood}
-                if dataset == "MedicalDomain":
-                    text, answer = row
-                    if split == "test":
-                        answer = ""
-                    yield id_, {"text": text,
-                                "answer": answer}
-
+                if split == "test":
+                    row += [""] * num_labels
+                return {name: value for name, value in zip(column_names, row)}