Datasets:

ICLR2024

/

ICLR2024-papers

like 2

Follow

ICLR2024 6

Poster

Deep learning (DL) models for tabular data problems (e.g. classification, regression) are currently receiving increasingly more attention from researchers.However, despite the recent efforts, the non-DL algorithms based on gradient-boosted decision trees (GBDT) remain a strong go-to solution for these problems.One of the research directions aimed at improving the position of tabular DL involves designing so-called retrieval-augmented models.For a target object, such models retrieve other objects (e.g. the nearest neighbors) from the available training data and use their features and labels to make a better prediction.In this work, we present TabR -- essentially, a feed-forward network with a custom k-Nearest-Neighbors-like component in the middle.On a set of public benchmarks with datasets up to several million objects, TabR marks a big step forward for tabular DL: it demonstrates the best average performance among tabular DL models, becomes the new state-of-the-art on several datasets, and even outperforms GBDT models on the recently proposed "GBDT-friendly" benchmark (see Figure 1).Among the important findings and technical details powering TabR, the main ones lie in the attention-like mechanism that is responsible for retrieving the nearest neighbors and extracting valuable signal from them.In addition to the higher performance, TabR is simple and significantly more efficient compared to prior retrieval-based tabular DL models.

TabR: Tabular Deep Learning Meets Nearest Neighbors

2307.14338

https://openreview.net/forum?id=rhgIgTSSxW

Poster

Generating a set of text is a common challenge for many NLP applications, for example, automatically providing multiple keyphrases for a document to facilitate user reading. Existing generative models use a sequential decoder which generates a single sequence successively, and the set generation problem is converted to sequence generation via concatenating multiple texts into a long text sequence. However, the elements of a set are unordered, which makes this scheme suffer from biased or conflicting training signals. In this paper, we propose a novel branching decoder. It can generate a dynamic number of tokens at each time-step and branch multiple generation paths. In particular, paths are generated individually so that no order dependence is required. Moreover, multiple paths can be generated in parallel which greatly reduces inference time. Experiments on three keyphrase generation datasets demonstrate that our branching decoder is more effective and efficient than the existing sequential decoder.

A Branching Decoder for Set Generation

https://openreview.net/forum?id=riNuqYiD66

Poster

Language model (LM) prompting—a popular paradigm for solving NLP tasks—has been shown to be susceptible to miscalibration and brittleness to slight prompt variations, caused by its discriminative prompting approach, i.e., predicting the label given the input. To address these issues, we propose Gen-Z—a generative prompting framework for zero-shot text classification. GEN-Z is generative, as it measures the LM likelihood of input text, conditioned on natural language descriptions of labels. The framework is multivariate, as label descriptions allow us to seamlessly integrate additional contextual information about the labels to improve task performance. On various standard classification benchmarks, with six open-source LM families, we show that zero-shot classification with simple contextualization of the data source of the evaluation set consistently outperforms both zero-shot and few-shot baselines while improving robustness to prompt variations. Further, our approach enables personalizing classification in a zero-shot manner by incorporating author, subject, or reader information in the label descriptions.

Gen-Z: Generative Zero-Shot Text Classification with Contextualized Label Descriptions

https://openreview.net/forum?id=rkplYfqUr0

Spotlight Poster

Unsupervised pretraining, which learns a useful representation using a large amount of unlabeled data to facilitate the learning of downstream tasks, is a critical component of modern large-scale machine learning systems. Despite its tremendous empirical success, the rigorous theoretical understanding of why unsupervised pretraining generally helps remains rather limited---most existing results are restricted to particular methods or approaches for unsupervised pretraining with specialized structural assumptions. This paper studies a generic framework,where the unsupervised representation learning task is specified by an abstract class of latent variable models $\Phi$ and the downstream task is specified by a class of prediction functions $\Psi$. We consider a natural approach of using Maximum Likelihood Estimation (MLE) for unsupervised pretraining and Empirical Risk Minimization (ERM) for learning downstream tasks. We prove that, under a mild ``informative'' condition, our algorithm achieves an excess risk of $\\tilde{\\mathcal{O}}(\sqrt{\mathcal{C}\_\Phi/m} + \sqrt{\mathcal{C}\_\Psi/n})$ for downstream tasks, where $\mathcal{C}\_\Phi, \mathcal{C}\_\Psi$ are complexity measures of function classes $\Phi, \Psi$, and $m, n$ are the number of unlabeled and labeled data respectively. Comparing to the baseline of $\tilde{\mathcal{O}}(\sqrt{\mathcal{C}\_{\Phi \circ \Psi}/n})$ achieved by performing supervised learning using only the labeled data, our result rigorously shows the benefit of unsupervised pretraining when $m \gg n$ and $\mathcal{C}\_{\Phi\circ \Psi} > \mathcal{C}\_\Psi$. This paper further shows that our generic framework covers a wide range of approaches for unsupervised pretraining, including factor models, Gaussian mixture models, and contrastive learning.

On the Provable Advantage of Unsupervised Pretraining

2303.01566

https://openreview.net/forum?id=rmXXKxQpOR

Spotlight Poster

What is the best paradigm to recognize objects---discriminative inference (fast but potentially prone to shortcut learning) or using a generative model (slow but potentially more robust)? We build on recent advances in generative modeling that turn text-to-image models into classifiers. This allows us to study their behavior and to compare them against discriminative models and human psychophysical data.We report four intriguing emergent properties of generative classifiers: they show a record-breaking human-like shape bias (99% for Imagen), near human-level out-of-distribution accuracy, state-of-the-art alignment with human classification errors, and they understand certain perceptual illusions. Our results indicate that while the current dominant paradigm for modeling human object recognition is discriminative inference, zero-shot generative models approximate human object recognition data surprisingly well.

Intriguing Properties of Generative Classifiers

2309.16779

https://openreview.net/forum?id=rmg0qMKYRQ

Poster

Previous literature on policy diversity in reinforcement learning (RL) either focuses on the online setting or ignores the policy performance. In contrast, offline RL, which aims to learn high-quality policies from batched data, has yet to fully leverage the intrinsic diversity of the offline dataset. Addressing this dichotomy and aiming to balance quality and diversity poses a significant challenge to extant methodologies. This paper introduces a novel approach, termed Stylized Offline RL (SORL), which is designed to extract high-performing, stylistically diverse policies from a dataset characterized by distinct behavioral patterns. Drawing inspiration from the venerable Expectation-Maximization (EM) algorithm, SORL innovatively alternates between policy learning and trajectory clustering, a mechanism that promotes policy diversification. To further augment policy performance, we introduce advantage-weighted style learning into the SORL framework. Experimental evaluations across multiple environments demonstrate the significant superiority of SORL over previous methods in extracting high-quality policies with diverse behaviors. A case in point is that SORL successfully learns strong policies with markedly distinct playing patterns from a real-world human dataset of a popular basketball video game "Dunk City Dynasty."

Stylized Offline Reinforcement Learning: Extracting Diverse High-Quality Behaviors from Heterogeneous Datasets

https://openreview.net/forum?id=rnHNDihrIT

Poster

The demand for transparency in healthcare and finance has led to interpretable machine learning (IML) models, notably the concept bottleneck models (CBMs), valued for its potential in performance and insights into deep neural networks. However, CBM's reliance on manually annotated data poses challenges. Label-free CBM has emerged to address this, but they remain unstable, affecting their faithfulness as explanatory tools. To address this inherent instability issue, we introduce a formal definition for an alternative concept called the Faithful Vision-Language Concept (FVLC) models. We present a methodology for constructing an FVLC that satisfies four critical properties. Our extensive experimentation, conducted on four benchmark datasets using Label-free CBM model architectures, demonstrates that our FVLC outperforms other baselines in terms of stability against input and concept set perturbations. Our approach incurs minimal accuracy degradation compared to the vanilla CBM, making it a promising solution for reliable and faithful model interpretation.

Faithful Vision-Language Interpretation via Concept Bottleneck Models

https://openreview.net/forum?id=rp0EdI8X4e

Poster

Speculative decoding~(SD) accelerates large language model inference by employing a faster {\em draft} model for generating multiple tokens, which are then verified in parallel by the larger {\em target} model, resulting in the text generated according to the target model distribution. However, identifying a compact draft model that is well-aligned with the target model is challenging. To tackle this issue, we propose {\em DistillSpec} that uses knowledge distillation to better align the draft model with the target model, before applying SD. DistillSpec makes two key design choices, which we demonstrate via systematic study to be crucial to improve the draft and target alignment: utilizing \emph{on-policy} data generation from the draft model, and \emph{tailoring the divergence function} to the task and decoding strategy. Notably, DistillSpec yields impressive $10 - 45\%$ speedups over standard SD on a range of standard benchmarks, using both greedy and non-greedy sampling. Furthermore, we combine DistillSpec with lossy SD to achieve fine-grained control over the latency vs. task performance trade-off. Finally, in practical scenarios with models of varying sizes, first using distillation to boost the performance of the target model and then applying DistillSpec to train a well-aligned draft model can reduce decoding latency by $6 - 10\times$ with minimal performance drop, compared to standard decoding without distillation.

DistillSpec: Improving Speculative Decoding via Knowledge Distillation

2310.08461

https://openreview.net/forum?id=rsY6J3ZaTF

Poster

We introduce a principled way of computing the Wasserstein distance between two distributions in a federated manner.Namely, we show how to estimate the Wasserstein distance between two samples stored and kept on different devices/clients whilst a central entity/server orchestrates the computations (again, without having access to the samples). To achieve this feat, we take advantage of the geometric properties of the Wasserstein distance -- in particular, the triangle inequality -- and that of the associated {\em geodesics}: our algorithm, FedWad (for Federated Wasserstein Distance), iteratively approximates the Wasserstein distance by manipulating and exchanging distributions from the space of geodesics in lieu of the input samples. In addition to establishing the convergence properties of FedWad, we provide empirical results on federated coresets and federate optimal transport dataset distance, that we respectively exploit for building a novel federated model and for boosting performance of popular federated learning algorithms.

Federated Wasserstein Distance

2310.01973

https://openreview.net/forum?id=rsg1mvUahT

Poster

Recently, multimodal contrastive learning (MMCL) approaches, such as CLIP \citep{radford2021learning}, have achieved a remarkable success in learning representations that are robust against distribution shift and generalize to new domains. Despite the empirical success, the mechanism behind learning such generalizable representations is not understood. In this work, we rigorously analyze this problem and uncover two mechanisms behind MMCL's robustness: \emph{intra-class contrasting}, which allows the model to learn features with a high variance, and \emph{inter-class feature sharing}, where annotated details in one class help learning other classes better. Both mechanisms prevent spurious features that are over-represented in the training data to overshadow the generalizable core features. This yields superior zero-shot classification accuracy under distribution shift. Furthermore, we theoretically demonstrate the benefits of using rich captions on robustness and explore the effect of annotating different types of details in the captions. We validate our theoretical findings through experiments, including a well-designed synthetic experiment and an experiment involving training CLIP on MS COCO \citep{lin2014microsoft} and evaluating the model on variations of shifted ImageNet.

Understanding the Robustness of Multi-modal Contrastive Learning to Distribution Shift

2310.04971

https://openreview.net/forum?id=rtl4XnJYBh

Spotlight Poster

We present a novel variational framework for performing inference in (neural) stochastic differential equations (SDEs) driven by Markov-approximate fractional Brownian motion (fBM). SDEs offer a versatile tool for modeling real-world continuous-time dynamic systems with inherent noise and randomness. Combining SDEs with the powerful inference capabilities of variational methods, enables the learning of representative function distributions through stochastic gradient descent. However, conventional SDEs typically assume the underlying noise to follow a Brownian motion (BM), which hinders their ability to capture long-term dependencies. In contrast, fractional Brownian motion (fBM) extends BM to encompass non-Markovian dynamics, but existing methods for inferring fBM parameters are either computationally demanding or statistically inefficient. In this paper, building upon the Markov approximation of fBM, we derive the evidence lower bound essential for efficient variational inference of posterior path measures, drawing from the well-established field of stochastic analysis. Additionally, we provide a closed-form expression to determine optimal approximation coefficients. Furthermore, we propose the use of neural networks to learn the drift, diffusion and control terms within our variational posterior, leading to the variational training of neural-SDEs. In this framework, we also optimize the Hurst index, governing the nature of our fractional noise. Beyond validation on synthetic data, we contribute a novel architecture for variational latent video prediction,—an approach that, to the best of our knowledge, enables the first variational neural-SDE application to video perception.

Variational Inference for SDEs Driven by Fractional Noise

2310.12975

https://openreview.net/forum?id=rtx8B94JMS

Poster

Misclassification detection is an important problem in machine learning, as it allows for the identification of instances where the model's predictions are unreliable. However, conventional uncertainty measures such as Shannon entropy do not provide an effective way to infer the real uncertainty associated with the model's predictions. In this paper, we introduce a novel data-driven measure of relative uncertainty to an observer for misclassification detection. By learning patterns in the distribution of soft-predictions, our uncertainty measure can identify misclassified samples based on the predicted class probabilities. Interestingly, according to the proposed measure, soft-predictions that correspond to misclassified instances can carry a large amount of uncertainty, even though they may have low Shannon entropy. We demonstrate empirical improvements over multiple image classification tasks, outperforming state-of-the-art misclassification detection methods.

A Data-Driven Measure of Relative Uncertainty for Misclassification Detection

2306.01710

https://openreview.net/forum?id=ruGY8v10mK

Poster

Conformal prediction (CP) for regression can be challenging, especially when the output distribution is heteroscedastic, multimodal, or skewed. Some of the issues can be addressed by estimating a distribution over the output, but in reality, such approaches can be sensitive to estimation error and yield unstable intervals. Here, we circumvent the challenges by converting regression to a classification problem and then use CP for classification to obtain CP sets for regression. To preserve the ordering of the continuous-output space, we design a new loss function and present necessary modifications to the CP classification techniques. Empirical results on many benchmarks show that this simple approach gives surprisingly good results on many practical problems.

Conformal Prediction via Regression-as-Classification

2404.08168

https://openreview.net/forum?id=rulxyXjf46

Poster

Computational simulation of chemical and biological systems using *ab initio* molecular dynamics has been a challenge over decades. Researchers have attempted to address the problem with machine learning and fragmentation-based methods. However, the two approaches fail to give a satisfactory description of long-range and many-body interactions, respectively. Inspired by fragmentation-based methods, we propose the Long-Short-Range Message-Passing (LSR-MP) framework as a generalization of the existing equivariant graph neural networks (EGNNs) with the intent to incorporate long-range interactions efficiently and effectively. We apply the LSR-MP framework to the recently proposed ViSNet and demonstrate the state-of-the-art results with up to 40% MAE reduction for molecules in MD22 and Chignolin datasets. Consistent improvements to various EGNNs will also be discussed to illustrate the general applicability and robustness of our LSR-MP framework. The code for our experiments and trained model weights could be found at https://anonymous.4open.science/r/LSRM-760E/.

Long-Short-Range Message-Passing: A Physics-Informed Framework to Capture Non-Local Interaction for Scalable Molecular Dynamics Simulation

https://openreview.net/forum?id=rvDQtdMnOl

Poster

Bayesian deep learning counts on the quality of posterior distribution estimation. However, the posterior of deep neural networks is highly multi-modal in nature, with local modes exhibiting varying generalization performance. Given a practical budget, targeting at the original posterior can lead to suboptimal performance, as some samples may become trapped in "bad" modes and suffer from overfitting. Leveraging the observation that "good" modes with low generalization error often reside in flat basins of the energy landscape, we propose to bias sampling on the posterior toward these flat regions. Specifically, we introduce an auxiliary guiding variable, the stationary distribution of which resembles a smoothed posterior free from sharp modes, to lead the MCMC sampler to flat basins. By integrating this guiding variable with the model parameter, we create a simple joint distribution that enables efficient sampling with minimal computational overhead. We prove the convergence of our method and further show that it converges faster than several existing flatness-aware methods in the strongly convex setting. Empirical results demonstrate that our method can successfully sample from flat basins of the posterior, and outperforms all compared baselines on multiple benchmarks including classification, calibration, and out-of-distribution detection.

Entropy-MCMC: Sampling from Flat Basins with Ease

https://openreview.net/forum?id=oGNdBvymod

Spotlight Poster

Pre-training large models on vast amounts of web data has proven to be an effective approach for obtaining powerful, general models in several domains, including language and vision. However, this paradigm has not yet taken hold in deep reinforcement learning (RL). This gap is due to the fact that the most abundant form of embodied behavioral data on the web consists of videos, which do not include the action labels required by existing methods for training policies from offline data. We introduce Latent Action Policies from Observation (LAPO), a method to infer latent actions and, consequently, latent-action policies purely from action-free demonstrations. Our experiments on challenging procedurally-generated environments show that LAPO can act as an effective pre-training method to obtain RL policies that can then be rapidly fine-tuned to expert-level performance. Our approach serves as a key stepping stone to enabling the pre-training of powerful, generalist RL models on the vast amounts of action-free demonstrations readily available on the web.

Learning to Act without Actions

2312.10812

https://openreview.net/forum?id=rvUq3cxpDF

Spotlight Poster

The rise in internet usage has led to the generation of massive amounts of data, resulting in the adoption of various supervised and semi-supervised machine learning algorithms, which can effectively utilize the colossal amount of data to train models. However, before deploying these models in the real world, these must be strictly evaluated on performance measures like worst-case recall and satisfy constraints such as fairness. We find that current state-of-the-art empirical techniques offer sub-optimal performance on these practical, non-decomposable performance objectives. On the other hand, the theoretical techniques necessitate training a new model from scratch for each performance objective. To bridge the gap, we propose SelMix, a selective mixup-based inexpensive fine-tuning technique for pre-trained models, to optimize for the desired objective. The core idea of our framework is to determine a sampling distribution to perform a mixup of features between samples from particular classes such that it optimizes the given objective. We comprehensively evaluate our technique against the existing empirical and theoretically principled methods on standard benchmark datasets for imbalanced classification. We find that proposed SelMix fine-tuning significantly improves the performance for various practical non-decomposable objectives across benchmarks.

Selective Mixup Fine-Tuning for Optimizing Non-Decomposable Objectives

2403.18301

https://openreview.net/forum?id=rxVBKhyfSo

Poster

The ability to engineer novel proteins with higher fitness for a desired property would be revolutionary for biotechnology and medicine. Modeling the combinatorially large space of sequences is infeasible; prior methods often constrain optimization to a small mutational radius, but this drastically limits the design space. Instead of heuristics, we propose smoothing the fitness landscape to facilitate protein optimization. First, we formulate protein fitness as a graph signal then use Tikunov regularization to smooth the fitness landscape. We find optimizing in this smoothed landscape leads to improved performance across multiple methods in the GFP and AAV benchmarks. Second, we achieve state-of-the-art results utilizing discrete energy-based models and MCMC in the smoothed landscape. Our method, called Gibbs sampling with Graph-based Smoothing (GGS), demonstrates a unique ability to achieve 2.5 fold fitness improvement (with in-silico evaluation) over its training set. GGS demonstrates potential to optimize proteins in the limited data regime. Code: https://github.com/kirjner/GGS

Improving protein optimization with smoothed fitness landscapes

2307.00494

https://openreview.net/forum?id=rxlF2Zv8x0

Spotlight Poster

The visual examination of tissue biopsy sections is fundamental for cancer diagnosis, with pathologists analyzing sections at multiple magnifications to discern tumor cells and their subtypes. However, existing attention-based multiple instance learning (MIL) models, used for analyzing Whole Slide Images (WSIs) in cancer diagnostics, often overlook the contextual information of tumor and neighboring tiles, leading to misclassifications. To address this, we propose the Context-Aware Multiple Instance Learning (CAMIL) architecture. CAMIL incorporates neighbor-constrained attention to consider dependencies among tiles within a WSI and integrates contextual constraints as prior knowledge into the MIL model. We evaluated CAMIL on subtyping non-small cell lung cancer (TCGA-NSCLC) and detecting lymph node (CAMELYON16) metastasis, achieving test AUCs of 0.959\% and 0.975\%, respectively, outperforming other state-of-the-art methods. Additionally, CAMIL enhances model interpretability by identifying regions of high diagnostic value

CAMIL: Context-Aware Multiple Instance Learning for Cancer Detection and Subtyping in Whole Slide Images

2305.05314

https://openreview.net/forum?id=rzBskAEmoc

Poster

In this paper, we tackle the problem of 3D reconstruction of dynamic scenes from multi-view videos. Previous works attempt to model the motion of 3D points in space, which either constrains them to handle a single articulated object or requires extra efforts to handle topology changes. By contrast, we propose to directly estimate the change of Signed Distance Function (SDF), namely SDF flow, of the dynamic scene. We show that the SDF flow captures the evolution of the scene surface and handles topology changes naturally. We further derive the mathematical relation between the SDF flow and the scene flow, which allows us to calculate the scene flow from the SDF flow analytically by solving linear equations. Our experiments on real-world multi-view video datasets show that our reconstructions are better than those of the state-of-the-art methods.

Neural SDF Flow for 3D Reconstruction of Dynamic Scenes

https://openreview.net/forum?id=rzF0R6GOd4

Poster

Given a pretrained encoder-based language model, how can we accurately compress it without retraining? Retraining-free structured pruning algorithms are crucial in pretrained language model compression due to their significantly reduced pruning cost and capability to prune large language models. However, existing retraining-free algorithms encounter severe accuracy degradation, as they fail to handle pruning errors, especially at high compression rates. In this paper, we propose KPrune (Knowledge-preserving pruning), an accurate retraining-free structured pruning algorithm for pretrained encoder-based language models.KPrune focuses on preserving the useful knowledge of the pretrained model to minimize pruning errors through a carefully designed iterative pruning process composed of knowledge measurement, knowledge-preserving mask search, and knowledge-preserving weight-tuning. As a result, KPrune shows significant accuracy improvements up to 58.02%p higher F1 score compared to existing retraining-free pruning algorithms under a high compression rate of 80% on the SQuAD benchmark without any retraining process.

Accurate Retraining-free Pruning for Pretrained Encoder-based Language Models

2308.03449

https://openreview.net/forum?id=s2NjWfaYdZ

Poster

We present a novel defense, against backdoor attacks on Deep Neural Networks (DNNs), wherein adversaries covertly implant malicious behaviors (backdoors) into DNNs. Our defense falls within the category of post-development defenses that operate independently of how the model was generated. The proposed defense is built upon a novel reverse engineering approach that can directly extract **backdoor functionality** of a given backdoored model to a *backdoor expert* model. The approach is straightforward --- finetuning the backdoored model over a small set of intentionally mislabeled clean samples, such that it unlearns the normal functionality while still preserving the backdoor functionality, and thus resulting in a model~(dubbed a backdoor expert model) that can only recognize backdoor inputs. Based on the extracted backdoor expert model, we show the feasibility of devising highly accurate backdoor input detectors that filter out the backdoor inputs during model inference. Further augmented by an ensemble strategy with a finetuned auxiliary model, our defense, **BaDExpert** (**Ba**ckdoor Input **D**etection with Backdoor **Expert**), effectively mitigates 17 SOTA backdoor attacks while minimally impacting clean utility. The effectiveness of BaDExpert has been verified on multiple datasets (CIFAR10, GTSRB and ImageNet) across various model architectures (ResNet, VGG, MobileNetV2 and Vision Transformer).

BaDExpert: Extracting Backdoor Functionality for Accurate Backdoor Input Detection

2308.12439

https://openreview.net/forum?id=s56xikpD92

Poster

The standard active learning setting assumes a willing labeler, who provides labels on informative examples to speed up learning. However, if the labeler wishes to be compensated for as many labels as possible before learning finishes, the labeler may benefit from actually slowing down learning. This incentive arises for instance if the labeler is to be replaced by the ML model, once it is learned. In this paper, we initiate the study of learning from a strategic labeler, who selectively abstains from labeling to slow down learning. We first prove that strategic abstention can prolong learning, and propose novel complexity measures to analyze the query cost of the learning game. Next, we develop a near-optimal deterministic algorithm, prove its robustness to strategic labeling, and contrast it with other active learning algorithms. We also provide extensions that encompass other learning setups/goals. Finally, we characterize the query cost of multi-task active learning, with and without abstention. Our first exploration of strategic labeling aims to add to our theoretical understanding of the imitative nature of ML in human-AI interaction.

The Human-AI Substitution game: active learning from a strategic labeler

https://openreview.net/forum?id=s5hSp7EdL3

Poster

Gradient inversion attacks aim to reconstruct local training data from intermediate gradients exposed in the federated learning framework. Despite successful attacks, all previous methods, starting from reconstructing a single data point and then relaxing the single-image limit to batch level, only work well under hard label constraints. Even for single-image reconstruction, we still lack an analysis-based algorithm to recover augmented soft labels. In this work, we change the focus from enlarging batchsize to investigating the hard label constraints, considering a more realistic circumstance where label smoothing and mixup techniques are used in the training process. In particular, we are the first to initialize a novel algorithm to simultaneously recover the ground-truth augmented label and the input feature of the last fully-connected layer from single-input gradients, and provide a necessary condition for any analytical-based label recovery methods. Extensive experiments testify to the label recovery accuracy, as well as the benefits to the following image reconstruction. We believe soft labels in classification tasks are worth further attention in gradient inversion attacks.

Towards Eliminating Hard Label Constraints in Gradient Inversion Attacks

2402.03124

https://openreview.net/forum?id=s8cMuxI5gu

Poster

Training and deploying machine learning models that meet fairness criteria for protected groups are fundamental in modern artificial intelligence. While numerous constraints and regularization terms have been proposed in the literature to promote fairness in machine learning tasks, most of these methods are not amenable to stochastic optimization due to the complex and nonlinear structure of constraints and regularizers. Here, the term "stochastic" refers to the ability of the algorithm to work with small mini-batches of data. Motivated by the limitation of existing literature, this paper presents a unified stochastic optimization framework for fair empirical risk minimization based on $f$-divergence measures ($f$-FERM). The proposed stochastic algorithm enjoys theoretical convergence guarantees. In addition, our experiments demonstrate the superiority of fairness-accuracy tradeoffs offered by $f$-FERM for almost all batch sizes (ranging from full-batch to batch size of one). Moreover, we show that our framework can be extended to the case where there is a distribution shift from training to the test data. Our extension is based on a distributionally robust optimization reformulation of $f$-FERM objective under $\ell_p$ norms as uncertainty sets. Again in this distributionally robust setting, $f$-FERM not only enjoys theoretical convergence guarantees but also outperforms other baselines in the literature in the tasks involving distribution shifts. An efficient stochastic implementation of $f$-FERM is publicly available.

f-FERM: A Scalable Framework for Robust Fair Empirical Risk Minimization

https://openreview.net/forum?id=s90VIdza2K

Poster

Recent advancements in diffusion models have significantly enhanced the data synthesis with 2D control. Yet, precise 3D control in street view generation, crucial for 3D perception tasks, remains elusive. Specifically, utilizing Bird's-Eye View (BEV) as the primary condition often leads to challenges in geometry control (e.g., height), affecting the representation of object shapes, occlusion patterns, and road surface elevations, all of which are essential to perception data synthesis, especially for 3D object detection tasks. In this paper, we introduce MagicDrive, a novel street view generation framework offering diverse 3D geometry controls, including camera poses, road maps, and 3D bounding boxes, together with textual descriptions, achieved through tailored encoding strategies. Besides, our design incorporates a cross-view attention module, ensuring consistency across multiple camera views. With MagicDrive, we achieve high-fidelity street-view synthesis that captures nuanced 3D geometry and various scene descriptions, enhancing tasks like BEV segmentation and 3D object detection. Project Website: https://magic-drive.github.io/

MagicDrive: Street View Generation with Diverse 3D Geometry Control

2310.02601

https://openreview.net/forum?id=sBQwvucduK

Poster

Universal domain adaptation aims to align the classes and reduce the feature gap between the same category of the source and target domains. The target private category is set as the unknown class during the adaptation process, as it is not included in the source domain. However, most existing methods overlook the intra-class structure within a category, especially in cases where there exists significant concept shift between the samples belonging to the same category. When samples with large concept shift are forced to be pushed together, it may negatively affect the adaptation performance. Moreover, from the interpretability aspect, it is unreasonable to align visual features with significant differences, such as fighter jets and civil aircraft, into the same category. Unfortunately, due to such semantic ambiguity and annotation cost, categories are not always classified in detail, making it difficult for the model to perform precise adaptation. To address these issues, we propose a novel Memory-Assisted Sub-Prototype Mining (MemSPM) method that can learn the differences between samples belonging to the same category and mine sub-classes when there exists significant concept shift between them. By doing so, our model learns a more reasonable feature space that enhances the transferability and reflects the inherent differences among samples annotated as the same category. We evaluate the effectiveness of our MemSPM method over multiple scenarios, including UniDA, OSDA, and PDA. Our method achieves state-of-the-art performance on four benchmarks in most cases.

Memory-Assisted Sub-Prototype Mining for Universal Domain Adaptation

2310.05453

https://openreview.net/forum?id=sGVmr7KHfn

Poster

The current advancements in open domain text generation have been spearheaded by Transformer-based large language models. Leveraging efficient parallelization and vast training datasets, these models achieve unparalleled text generation capabilities. Even so, current models are known to suffer from deficiencies such as repetitive texts, looping issues, and lack of robustness. While adversarial training through generative adversarial networks (GAN) is a proposed solution, earlier research in this direction has predominantly focused on older architectures, or narrow tasks. As a result, this approach is not yet compatible with modern language models for open-ended text generation, leading to diminished interest within the broader research community. We propose a computationally efficient GAN approach for sequential data that utilizes the parallelization capabilities of Transformer models. Our method revolves around generating multiple branching sequences from each training sample, while also incorporating the typical next-step prediction loss on the original data. In this way, we achieve a dense reward and loss signal for both the generator and the discriminator, resulting in a stable training dynamic. We apply our training method to pre-trained language models, using data from their original training set but less than 0.01% of the available data. A comprehensive human evaluation shows that our method significantly improves the quality of texts generated by the model while avoiding the previously reported sparsity problems of GAN approaches. Even our smaller models outperform larger original baseline models with more than 16 times the number of parameters. Finally, we corroborate previous claims that perplexity on held-out data is not a sufficient metric for measuring the quality of generated texts.

Branch-GAN: Improving Text Generation with (not so) Large Language Models

https://openreview.net/forum?id=sHEJJmzBIN

Poster

Deep representation learning methods struggle with continual learning, suffering from both catastrophic forgetting of useful units and loss of plasticity, often due to rigid and unuseful units. While many methods address these two issues separately, only a few currently deal with both simultaneously. In this paper, we introduce Utility-based Perturbed Gradient Descent (UPGD) as a novel approach for the continual learning of representations. UPGD combines gradient updates with perturbations, where it applies smaller modifications to more useful units, protecting them from forgetting, and larger modifications to less useful units, rejuvenating their plasticity. We adopt the challenging setup of streaming learning as the testing ground and design continual learning problems with hundreds of non-stationarities and unknown task boundaries. We show that many existing methods suffer from at least one of the issues, predominantly manifested by their decreasing accuracy over tasks. On the other hand, UPGD continues to improve performance and surpasses or is competitive with all methods in all problems, being among the few methods demonstrably capable of addressing both issues.

Addressing Loss of Plasticity and Catastrophic Forgetting in Continual Learning

2404.00781

https://openreview.net/forum?id=sKPzAXoylB

Poster

Federated learning is an emerging technology for training machine learning models across decentralized data sources without sharing data. Vertical federated learning, also known as feature-based federated learning, applies to scenarios where data sources have the same sample IDs but different feature sets. To ensure fairness among data owners, it is critical to objectively assess the contributions from different data sources and compensate the corresponding data owners accordingly. The Shapley value is a provably fair contribution valuation metric originating from cooperative game theory. However, its straight-forward computation requires extensively retraining a model on each potential combination of data sources, leading to prohibitively high communication and computation overheads due to multiple rounds of federated learning. To tackle this challenge, we propose a contribution valuation metric called vertical federated Shapley value (VerFedSV) based on the classic Shapley value. We show that VerFedSV not only satisfies many desirable properties of fairness but is also efficient to compute. Moreover, VerFedSV can be adapted to both synchronous and asynchronous vertical federated learning algorithms. Both theoretical analysis and extensive experimental results demonstrate the fairness, efficiency, adaptability, and effectiveness of VerFedSV.

Fair and Efficient Contribution Valuation for Vertical Federated Learning

2201.02658

https://openreview.net/forum?id=sLQb8q0sUi

Spotlight Poster

Spatio-temporal (ST) prediction plays a pivotal role in earth sciences, such as meteorological prediction, urban computing. Adequate high-quality data, coupled with deep models capable of inference, are both indispensable and prerequisite for achieving meaningful results. However, the sparsity of data and the high costs associated with deploying sensors lead to significant data imbalances. Models that are overly tailored and lack causal relationships further compromise the generalizabilities of inference methods. Towards this end, we first establish a causal concept for ST predictions, named NuwaDynamics, which targets to identify causal regions in data and endow model with causal reasoning ability in a two-stage process. Concretely, we initially leverage upstream self-supervision to discern causal important patches, imbuing the model with generalized information and conducting informed interventions on complementary trivial patches to extrapolate potential test distributions. This phase is referred to as the discovery step. Advancing beyond discovery step, we transfer the data to downstream tasks for targeted ST objectives, aiding the model in recognizing a broader potential distribution and fostering its causal perceptual capabilities (refer as Update step). Our concept aligns seamlessly with the contemporary backdoor adjustment mechanism in causality theory. Extensive experiments on six real-world ST benchmarks showcase that models can gain outcomes upon the integration of the NuwaDynamics concept. NuwaDynamics also can significantly benefit a wide range of changeable ST tasks like extreme weather and long temporal step super-resolution predictions.

NuwaDynamics: Discovering and Updating in Causal Spatio-Temporal Modeling

https://openreview.net/forum?id=sLdVl0q68X

Poster

Mixture models arise in many regression problems, but most methods have seen limited adoption partly due to these algorithms' highly-tailored and model-specific nature. On the other hand, transformers are flexible, neural sequence models that present the intriguing possibility of providing general-purpose prediction methods, even in this mixture setting. In this work, we investigate the hypothesis that transformers can learn an optimal predictor for mixtures of regressions. We construct a generative process for a mixture of linear regressions for which the decision-theoretic optimal procedure is given by data-driven exponential weights on a finite set of parameters. We observe that transformers achieve low mean-squared error on data generated via this process. By probing the transformer's output at inference time, we also show that transformers typically make predictions that are close to the optimal predictor. Our experiments also demonstrate that transformers can learn mixtures of regressions in a sample-efficient fashion and are somewhat robust to distribution shifts. We complement our experimental observations by proving constructively that the decision-theoretic optimal procedure is indeed implementable by a transformer.

Transformers can optimally learn regression mixture models

2311.08362

https://openreview.net/forum?id=sLkj91HIZU

Poster

Image-to-image (I2I) translation is vital in computer vision tasks like style transfer and domain adaptation. While recent advances in GAN have enabled high-quality sample generation, real-world challenges such as noise and distortion remain significant obstacles. Although Gaussian noise injection during training has been utilized, its theoretical underpinnings have been unclear. This work provides a robust theoretical framework elucidating the role of Gaussian noise injection in I2I translation models. We address critical questions on the influence of noise variance on distribution divergence, resilience to unseen noise types, and optimal noise intensity selection. Our contributions include connecting $f$-divergence and score matching, unveiling insights into the impact of Gaussian noise on aligning probability distributions, and demonstrating generalized robustness implications. We also explore choosing an optimal training noise level for consistent performance in noisy environments. Extensive experiments validate our theoretical findings, showing substantial improvements over various I2I baseline models in noisy settings. Our research rigorously grounds Gaussian noise injection for I2I translation, offering a sophisticated theoretical understanding beyond heuristic applications.

On the Analysis of GAN-based Image-to-Image Translation with Gaussian Noise Injection

https://openreview.net/forum?id=sLregLuXpn

Spotlight Poster

Pruning is one of the mainstream methods to compress over-parameterized neural networks, resulting in significant practical benefits. Recently, another line of work has explored the direction of fusion, i.e. merging, independently trained neural networks. Here, we seek to marry the two approaches in a bid to combine their advantages into a single approach, which we term `Intra-Fusion'. Specifically, we implicitly utilize the pruning criteria to result in more informed fusion. Agnostic to the choice of a specific neuron-importance metric, Intra-Fusion can typically prune an additional considerable amount of the parameters while retaining the same accuracy as the standard pruning approach. Additionally, we explore how fusion can be added to the pruning process to significantly decrease the training time while maintaining competitive performance. We benchmark our results for various networks on commonly used datasets such as CIFAR10, CIFAR100, and ImageNet. More broadly, we hope that the proposed approach invigorates exploration into a fresh alternative to the predominant compression approaches.

Towards Meta-Pruning via Optimal Transport

2402.07839

https://openreview.net/forum?id=sMoifbuxjB

Poster

In this work, we propose a novel link prediction model and further boost it by studying graph incompleteness. First, We introduce MPNN-then-SF, an innovative architecture leveraging structural feature (SF) to guide MPNN's representation pooling, with its implementation, namely Neural Common Neighbor (NCN). NCN exhibits superior expressiveness and scalability compared with existing models, which can be classified into two categories: SF-then-MPNN, augmenting MPNN's input with SF, and SF-and-MPNN, decoupling SF and MPNN. Second, we investigate the impact of graph incompleteness---the phenomenon that some links are unobserved in the input graph---on SF, like the common neighbor. Through dataset visualization, we observe that incompleteness reduces common neighbors and induces distribution shifts, significantly affecting model performance. To address this issue, we propose to use a link prediction model to complete the common neighbor structure. Combining this method with NCN, we propose Neural Common Neighbor with Completion (NCNC). NCN and NCNC outperform recent strong baselines by large margins, and NCNC further surpasses state-of-the-art models in standard link prediction benchmarks.

Neural Common Neighbor with Completion for Link Prediction

2302.00890

https://openreview.net/forum?id=sNFLN3itAd

Poster

Generating realistic human motions with high framerate is an underexplored task, due to the varied framerates of training data, huge memory burden brought by high framerates and slow sampling speed of generative models. Recent advances make a compromise for training by downsampling high-framerate details away and discarding low-framerate samples, which suffer from severe information loss and restricted-framerate generation. In this paper, we found that the recent emerging paradigm of Implicit Neural Representations (INRs) that encode a signal into a continuous function can effectively tackle this challenging problem. To this end, we introduce NeRM, a generative model capable of taking advantage of varied-size data and capturing variational distribution of motions for high-framerate motion synthesis. By optimizing latent representation and a auto-decoder conditioned on temporal coordinates, NeRM learns continuous motion fields of sampled motion clips that ingeniously avoid explicit modeling of raw varied-size motions. This expressive latent representation is then used to learn a diffusion model that enables both unconditional and conditional generation of human motions. We demonstrate that our approach achieves competitive results with state-of-the-art methods, and can generate arbitrary framerate motions. Additionally, we show that NeRM is not only memory-friendly, but also highly efficient even when generating high-framerate motions.

NeRM: Learning Neural Representations for High-Framerate Human Motion Synthesis

https://openreview.net/forum?id=sOJriBlOFd

Poster

Integrals with discontinuous integrands are ubiquitous, arising from discrete structure in applications like topology optimization, graphics, and computational geometry. These integrals are often part of a forward model in an inverse problem where it is necessary to reason backwards about the parameters, ideally using gradient-based optimization. Monte Carlo methods are widely used to estimate the value of integrals, but this results in a non-differentiable approximation that is amenable to neither conventional automatic differentiation nor reparameterization-based gradient methods. This significantly disrupts efforts to integrate machine learning methods in areas that exhibit these discontinuities: physical simulation and robotics, design, graphics, and computational geometry. Although bespoke domain-specific techniques can handle special cases, a general methodology to wield automatic differentiation in these discrete contexts is wanting. We introduce a differentiable variant of the simple Monte Carlo estimator which samples line segments rather than points from the domain. We justify our estimator analytically as conditional Monte Carlo and demonstrate the diverse functionality of the method as applied to image stylization, topology optimization, and computational geometry.

Fiber Monte Carlo

https://openreview.net/forum?id=sP1tCl2QBk

Poster

In this paper, we present Consistent4D, a novel approach for generating 4D dynamic objects from uncalibrated monocular videos. Uniquely, we cast the 360-degree dynamic object reconstruction as a 4D generation problem, eliminating the need for tedious multi-view data collection and camera calibration. This is achieved by leveraging the object-level 3D-aware image diffusion model as the primary supervision signal for training dynamic Neural Radiance Fields (DyNeRF). Specifically, we propose a cascade DyNeRF to facilitate stable convergence and temporal continuity under the supervision signal which is discrete along the time axis. To achieve spatial and temporal consistency, we further introduce an interpolation-driven consistency loss. It is optimized by minimizing the L2 distance between rendered frames from DyNeRF and interpolated frames from a pre-trained video interpolation model. Extensive experiments show that our Consistent4D can perform competitively to prior art alternatives, opening up new possibilities for 4D dynamic object generation from monocular videos, whilst also demonstrating advantage for conventional text-to-3D generation tasks

Consistent4D: Consistent 360° Dynamic Object Generation from Monocular Video

2311.02848

https://openreview.net/forum?id=sPUrdFGepF

Poster

In this paper, we aim to adapt a model at test-time using a few unlabeled data to address distribution shifts. In this setting, extracting the domain knowledge from a limited amount of data is challenging. To improve such a process, it is crucial to utilize correlated information from pre-trained backbones and source domains. Previous studies fail to utilize recent foundation models with strong out-of-distribution generalization. Additionally, domain-centric designs are not flavored in their works. Furthermore, they employ the process of modelling source domains and the process of learning to adapt independently into disjoint training stages. In this work, we propose an approach on top of the pre-computed features of the foundation model. Specifically, we build a knowledge bank to learn the transferable knowledge from source domains. Conditioned on few-shot target data, we introduce a domain prompt generator to condense the knowledge bank into a domain-specific prompt. The domain prompt then directs the visual features towards a particular domain via a guidance module. Moreover, we propose a domain-aware contrastive loss and employ meta-learning to facilitate domain knowledge extraction. Extensive experiments are conducted to validate the domain knowledge extraction. The proposed method outperforms previous work significantly on 5 large-scale benchmarks including WILDS and DomainNet.

Adapting to Distribution Shift by Visual Domain Prompt Generation

https://openreview.net/forum?id=sSaN4gxuEf

Poster

Class-incremental learning is becoming more popular as it helps models widen their applicability while not forgetting what they already know. A trend in this area is to use a mixture-of-expert technique, where different models work together to solve the task. However, the experts are usually trained all at once using whole task data, which makes them all prone to forgetting and increasing computational burden. To address this limitation, we introduce a novel approach named SEED. SEED selects only one, the most optimal expert for a considered task, and uses data from this task to fine-tune only this expert. For this purpose, each expert represents each class with a Gaussian distribution, and the optimal expert is selected based on the similarity of those distributions. Consequently, SEED increases diversity and heterogeneity within the experts while maintaining the high stability of this ensemble method. The extensive experiments demonstrate that SEED achieves state-of-the-art performance in exemplar-free settings across various scenarios, showing the potential of expert diversification through data in continual learning.

Divide and not forget: Ensemble of selectively trained experts in Continual Learning

2401.10191

https://openreview.net/forum?id=sSyytcewxe

Poster

Self-supervised learning methods have shown promising results across a wide range of tasks in computer vision, natural language processing, and multimodal analysis. However, self-supervised approaches come with a notable limitation, dimensional collapse, where a model doesn't fully utilize its capacity to encode information optimally. Motivated by this, we propose ProSMin, a novel probabilistic self-supervised learning approach that leverages the power of probabilistic models to enhance representation quality and mitigate collapsing representations. Our proposed approach involves two neural networks; the online network and the target network, which collaborate and learn the diverse distribution of representations from each other through probabilistic knowledge distillation. The two networks are trained via our new loss function based on proper scoring rules. We provide a theoretical justification for ProSMin's convergence, demonstrating the strict propriety of its modified scoring rule. This insight validates the method's optimization process and contributes to its robustness and effectiveness in improving representation quality. We evaluate our probabilistic model on various downstream tasks, such as in-distribution generalization, out-of-distribution detection, dataset corruption, low-shot learning, and transfer learning. Our method achieves superior accuracy and calibration, outperforming the self-supervised baseline in a variety of experiments on large datasets such as ImageNet-O and ImageNet-C. ProSMin thus demonstrates its scalability and real-world applicability. The code is in the supplementary material.

Probabilistic Self-supervised Representation Learning via Scoring Rules Minimization

https://openreview.net/forum?id=skcTCdJz0f

Poster

Protein structure representation learning is the foundation for promising applications in drug discovery, protein design, and protein function prediction. However, there remains a need for a robust, standardised benchmark to track the progress of new and established methods with greater granularity and relevance to downstream applications. In this work, we introduce a comprehensive and open benchmark suite for evaluating protein structure representation learning methods.We provide several pre-training methods, downstream tasks and pre-training corpora comprised of both experimental and predicted structures, offering a balanced challenge to representation learning algorithms. These tasks enable the systematic evaluation of the quality of the learned embeddings, the structural and functional relationships captured, and their usefulness in downstream tasks. We benchmark state-of-the-art protein-specific and generic geometric Graph Neural Networks and the extent to which they benefit from different types of pre-training. We find that pre-training consistently improves the performance of both rotation-invariant and equivariant models, and that equivariant models seem to benefit even more from pre-training compared to invariant models.We aim to establish a common ground for the machine learning and computational biology communities to collaborate, compare, and advance protein structure representation learning. By providing a standardised and rigorous evaluation platform, we expect to accelerate the development of novel methodologies and improve our understanding of protein structures and their functions. The codebase incorporates several engineering contributions which considerably reduces the barrier to entry for pre-training and working with large structure-based datasets. Our benchmark is available at: https://anonymous.4open.science/r/ProteinWorkshop-B8F5/

Evaluating Representation Learning on the Protein Structure Universe

https://openreview.net/forum?id=sTYuRVrdK3

Poster

Machine learning models are often used to automate routine tasks. In settings where mistakes are costly, we can trade off accuracy for coverage by abstaining from making a prediction on instances for which the model is uncertain. In this work, we present a new approach to selective classification in deep learning with concepts. Our approach constructs a concept bottleneck model where the front-end model can make predictions given soft concepts and leverage concept confirmation to improve coverage and performance under abstention. We develop techniques to propagate uncertainty and identify concepts for confirmation. We evaluate our approach on real-world and synthetic datasets, showing that it can improve coverage while maintaining performance across a range of tasks.

Classification with Conceptual Safeguards

https://openreview.net/forum?id=t8cBsT9mcg

Poster

Large language models (LMs) are capable of generating free-text rationales to aid question answering. However, prior work 1) suggests that useful self-rationalization is emergent only at significant scales (e.g., 175B parameter GPT-3); and 2) focuses largely on downstream performance, ignoring the semantics of the rationales themselves, e.g., are they faithful, true, and helpful for humans? In this work, we enable small-scale LMs (∼200x smaller than GPT-3) to generate rationales that not only improve downstream task performance, but are also more plausible, consistent, and diverse, assessed both by automatic and human evaluation. Our method, MaRio (Multi-rewArd RatIOnalization), is a multi-reward conditioned self-rationalization algorithm that optimizes multiple distinct properties like plausibility, diversity and consistency. Results on three difficult question-answering datasets StrategyQA, QuaRel and OpenBookQA show that not only does MaRio improve task accuracy, but it also improves the self-rationalization quality of small LMs across the aforementioned axes better than a supervised fine-tuning (SFT) baseline. Extensive human evaluations confirm that MaRio rationales are preferred vs. SFT rationales, as well as qualitative improvements in plausibility and consistency.

Tailoring Self-Rationalizers with Multi-Reward Distillation

2311.02805

https://openreview.net/forum?id=t8eO0CiZJV

Poster

Given a prompt, language models produce a distribution over all possible next tokens; when the prompt is unknown, can we use this distributional information to recover the prompt? We consider the problem of anguage model inversion and show that next-token probabilities contain a surprising amount of information about the preceding text. Often we can recover the text in cases where it is hidden from the user, motivating a method for recovering unknown prompts given only the model's current distribution output. We consider a variety of model access scenarios, and show how even without predictions for every token in the vocabulary we can recover the probability vector through search and reconstruction of the input. On LLAMA-7B, our inversion method reconstructs prompts with a BLEU of $59$ and token-level F1 of $77$ and recovers $23\%$ of prompts exactly

Language Model Inversion

2311.13647

https://openreview.net/forum?id=t9dWHpGkPj

Poster

Deep Generative Models (DGMs) have been shown to be powerful tools for generating tabular data, as they have been increasingly able to capture the complex distributions that characterize them. However, to generate realistic synthetic data, it is often not enough to have a good approximation of their distribution, as it also requires compliance with constraints that encode essential background knowledge on the problem at hand. In this paper, we address this limitation and show how DGMs for tabular data can be transformed into Constrained Deep Generative Models (C-DGMs), whose generated samples are guaranteed to be compliant with the given constraints. This is achieved by automatically parsing the constraints and transforming them into a Constraint Layer (CL) seamlessly integrated with the DGM. Our extensive experimental analysis with various DGMs and tasks reveals that standard DGMs often violate constraints, some exceeding 95% non-compliance, while their corresponding C-DGMs are never non-compliant. Then, we quantitatively demonstrate that, at training time, C-DGMs are able to exploit the background knowledge expressed by the constraints to outperform their standard counterparts with up to 4.5% improvement in utility and detection. Further, we show how our CL does not necessarily need to be integrated at training time, as it can be also used as a guardrail at inference time, still producing some improvements in the overall performance of the models. Finally, we show that our CL does not hinder the sample generation time of the models.

How Realistic Is Your Synthetic Data? Constraining Deep Generative Models for Tabular Data

2402.04823

https://openreview.net/forum?id=tBROYsEz9G

Poster

Label imbalance and homophily-heterophily mixture are the fundamental problems encountered when applying Graph Neural Networks (GNNs) to Graph Fraud Detection (GFD) tasks. Existing GNN-based GFD models are designed to augment graph structure to accommodate the inductive bias of GNNs towards homophily, by excluding heterophilic neighbors during message passing. In our work, we argue that the key to applying GNNs for GFD is not to exclude but to {\em distinguish} neighbors with different labels. Grounded in this perspective, we introduce Partitioning Message Passing (PMP), an intuitive yet effective message passing paradigm expressly crafted for GFD. Specifically, in the neighbor aggregation stage of PMP, neighbors with different classes are aggregated with distinct node-specific aggregation functions. By this means, the center node can adaptively adjust the information aggregated from its heterophilic and homophilic neighbors, thus avoiding the model gradient being dominated by benign nodes which occupy the majority of the population. We theoretically establish a connection between the spatial formulation of PMP and spectral analysis to characterize that PMP operates an adaptive node-specific spectral graph filter, which demonstrates the capability of PMP to handle heterophily-homophily mixed graphs. Extensive experimental results show that PMP can significantly boost the performance on GFD tasks.

Partitioning Message Passing for Graph Fraud Detection

https://openreview.net/forum?id=tEgrUrUuwA

Poster

Offline reinforcement learning (RL) holds promise as a means to learn high-reward policies from a static dataset, without the need for further environment interactions. However, a key challenge in offline RL lies in effectively stitching portions of suboptimal trajectories from the static dataset while avoiding extrapolation errors arising due to a lack of support in the dataset. Existing approaches use conservative methods that are tricky to tune and struggle with multi-modal data or rely on noisy Monte Carlo return-to-go samples for reward conditioning. In this work, we propose a novel approach that leverages the expressiveness of latent diffusion to model in-support trajectory sequences as compressed latent skills. This facilitates learning a Q-function while avoiding extrapolation error via batch-constraining. The latent space is also expressive and gracefully copes with multi-modal data. We show that the learned temporally-abstract latent space encodes richer task-specific information for offline RL tasks as compared to raw state-actions. This improves credit assignment and facilitates faster reward propagation during Q-learning. Our method demonstrates state-of-the-art performance on the D4RL benchmarks, particularly excelling in long-horizon, sparse-reward tasks.

Reasoning with Latent Diffusion in Offline Reinforcement Learning

2309.06599

https://openreview.net/forum?id=tGQirjzddO

Spotlight Poster

Despite the widespread empirical success of ResNet, the generalization ability of deep ResNet is rarely explored beyond the lazy-training regime. In this work, we investigate ResNet in the limit of infinitely deep and wide neural networks, of which the gradient flow is described by a partial differential equation in the large-neural network limit, i.e., the \emph{mean-field} regime. To derive the generalization bounds under this setting, our analysis necessitates a shift from the conventional time-invariant Gram matrix employed in the lazy training regime to a time-variant, distribution-dependent version tailored to the mean-field regime. To this end, we provide a lower bound on the minimum eigenvalue of the Gram matrix under the mean-field regime. Besides, the traceability of the dynamic of Kullback-Leibler (KL) divergence is also required under the mean-field regime. We therefore establish the linear convergence of the empirical error and estimate the upper bound of the KL divergence over parameters distribution. The above two results are employed to build the uniform convergence for generalization bound via Rademacher complexity. Our results offer new insights into the generalization ability of deep ResNet beyond the lazy training regime and contribute to advancing the understanding of the fundamental properties of deep neural networks.

Generalization of Scaled Deep ResNets in the Mean-Field Regime

2403.09889

https://openreview.net/forum?id=tMzPZTvz2H

Poster

Diffusion models are the de-facto approach for generating high-quality images and videos, but learning high-dimensional models remains a formidable task due to computational and optimization challenges. Existing methods often resort to training cascaded models in pixel space, or using a downsampled latent space of a separately trained auto-encoder. In this paper, we introduce Matryoshka Diffusion (MDM), an end-to-end framework for high-resolution image and video synthesis. We propose a diffusion process that denoises inputs at multiple resolutions jointly and uses a NestedUNet architecture where features and parameters for small-scale inputs are nested within those of large scales. In addition, MDM enables a progressive training schedule from lower to higher resolutions, which leads to significant improvements in optimization for high-resolution generation. We demonstrate the effectiveness of our approach on various benchmarks, including class-conditioned image generation, high-resolution text-to-image, and text-to-video applications. Remarkably, we can train a single pixel-space model at resolutions of up to 1024x1024 pixels, demonstrating strong zero-shot generalization using the CC12M dataset, which contains only 12 million images.

Matryoshka Diffusion Models

2310.15111

https://openreview.net/forum?id=tOzCcDdH9O

Spotlight Poster

Generalizing to out-of-distribution (OOD) data or unseen domain, termed OOD generalization, still lacks appropriate theoretical guarantees. Canonical OOD bounds focus on different distance measurements between source and target domains but fail to consider the optimization property of the learned model. As empirically shown in recent work, sharpness of learned minimum influences OOD generalization. To bridge this gap between optimization and OOD generalization, we study the effect of sharpness on how a model tolerates data change in domain shift which is usually captured by "robustness" in generalization. In this paper, we give a rigorous connection between sharpness and robustness, which gives better OOD guarantees for robust algorithms. It also provides a theoretical backing for "flat minima leads to better OOD generalization". Overall, we propose a sharpness-based OOD generalization bound by taking robustness into consideration, resulting in a tighter bound than non-robust guarantees. Our findings are supported by the experiments on a ridge regression model, as well as the experiments on deep learning classification tasks.

Towards Robust Out-of-Distribution Generalization Bounds via Sharpness

2403.06392

https://openreview.net/forum?id=tPEwSYPtAC

Spotlight Poster

Designing reward functions is a longstanding challenge in reinforcement learning (RL); it requires specialized knowledge or domain data, leading to high costs for development. To address this, we introduce Text2Reward, a data-free framework that automates the generation and shaping of dense reward functions based on large language models (LLMs). Given a goal described in natural language, Text2Reward generates shaped dense reward functions as an executable program grounded in a compact representation of the environment. Unlike inverse RL and recent work that uses LLMs to write sparse reward codes or unshaped dense rewards with a constant function across timesteps, Text2Reward produces interpretable, free-form dense reward codes that cover a wide range of tasks, utilize existing packages, and allow iterative refinement with human feedback. We evaluate Text2Reward on two robotic manipulation benchmarks (ManiSkill2, MetaWorld) and two locomotion environments of MuJoCo. On 13 of the 17 manipulation tasks, policies trained with generated reward codes achieve similar or better task success rates and convergence speed than expert-written reward codes. For locomotion tasks, our method learns six novel locomotion behaviors with a success rate exceeding 94%. Furthermore, we show that the policies trained in the simulator with our method can be deployed in the real world. Finally, Text2Reward further improves the policies by refining their reward functions with human feedback. Video results are available at https://text-to-reward.github.io/

Text2Reward: Reward Shaping with Language Models for Reinforcement Learning

https://openreview.net/forum?id=tUM39YTRxH

Poster

Invariance learning algorithms that conditionally filter out domain-specific random variables as distractors, do so based only on the data semantics, and not the target domain under evaluation. We show that a provably optimal and sample-efficient way of learning conditional invariances is by relaxing the invariance criterion to be non-commutatively directed towards the target domain. Under domain asymmetry, i.e., when the target domain contains semantically relevant information absent in the source, the risk of the encoder $\varphi^*$ that is optimal on average across domains is strictly lower-bounded by the risk of the target-specific optimal encoder $\Phi^*_\tau$. We prove that non-commutativity steers the optimization towards $\Phi^*_\tau$ instead of $\varphi^*$, bringing the $\mathcal{H}$-divergence between domains down to zero, leading to a stricter bound on the target risk. Both our theory and experiments demonstrate that non-commutative invariance (NCI) can leverage source domain samples to meet the sample complexity needs of learning $\Phi^*_\tau$, surpassing SOTA invariance learning algorithms for domain adaptation, at times by over 2\%, approaching the performance of an oracle. Implementation is available at https://github.com/abhrac/nci.

Learning Conditional Invariances through Non-Commutativity

2402.11682

https://openreview.net/forum?id=tUVG9nGzgE

Spotlight Poster

In this paper, we investigate the problem of offline Preference-based Reinforcement Learning (PbRL) with human feedback where feedback is available in the form of preference between trajectory pairs rather than explicit rewards. Our proposed algorithm consists of two main steps: (1) estimate the implicit reward using Maximum Likelihood Estimation (MLE) with general function approximation from offline data and (2) solve a distributionally robust planning problem over a confidence set around the MLE. We consider the general reward setting where the reward can be defined over the whole trajectory and provide a novel guarantee that allows us to learn any target policy with a polynomial number of samples, as long as the target policy is covered by the offline data. This guarantee is the first of its kind with general function approximation. To measure the coverage of the target policy, we introduce a new single-policy concentrability coefficient, which can be upper bounded by the per-trajectory concentrability coefficient. We also establish lower bounds that highlight the necessity of such concentrability and the difference from standard RL, where state-action-wise rewards are directly observed. We further extend and analyze our algorithm when the feedback is given over action pairs.

Provable Offline Preference-Based Reinforcement Learning

2305.14816

https://openreview.net/forum?id=tVMPfEGT2w

Poster

Clinical predictive models often rely on patients’ electronic health records (EHR), but integrating medical knowledge to enhance predictions and decision-making is challenging. This is because personalized predictions require personalized knowledgegraphs (KGs), which are difficult to generate from patient EHR data. To address this, we propose GraphCare, an open-world framework that uses external KGs to improve EHR-based predictions. Our method extracts knowledge from large language models (LLMs) and external biomedical KGs to build patient-specific KGs, which are then used to train our proposed Bi-attention AugmenTed(BAT) graph neural network (GNN) for healthcare predictions. On two public datasets, MIMIC-III and MIMIC-IV, GraphCare surpasses baselines in four vital healthcare prediction tasks: mortality, readmission, length of stay (LOS), and drug recommendation. On MIMIC-III, it boosts AUROC by 17.6% and 6.6% for mortality and readmission, and F1-score by 7.9% and 10.8% for LOS and drug recommendation, respectively. Notably, GraphCare demonstrates a substantial edge in scenarios with limited data availability. Our findings highlight the potential of using external KGs in healthcare prediction tasks and demonstrate the promise of GraphCare in generating personalized KGs for promoting personalized medicine.

GraphCare: Enhancing Healthcare Predictions with Personalized Knowledge Graphs

2305.12788

https://openreview.net/forum?id=tVTN7Zs0ml

Spotlight Poster

Learning to reject (L2R) is a classical problem where one seeks a classifier capable of abstaining on low-confidence samples. Most prior work on L2R has focused on minimizing the standard misclassification error. However, in many real-world applications, the label distribution is highly imbalanced, necessitating alternate evaluation metrics such as the balanced error or the worst-group error that enforce equitable performance across both the head and tail classes. In this paper, we establish that traditional L2R methods can be grossly sub-optimal for such metrics, and show that this is due to an intricate dependence in the objective between the label costs and the rejector. We then derive the form of the Bayes-optimal classifier and rejector for the balanced error, propose a novel plug-in approach to mimic this solution, and extend our results to general evaluation metrics. Through experiments on benchmark image classification tasks, we show that our approach yields better trade-offs in both the balanced and worst-group error compared to L2R baselines.

Learning to Reject for Balanced Error and Beyond

https://openreview.net/forum?id=ta26LtNq2r

Spotlight Poster

Large language models (LLMs) can perform impressive feats with in-context learning or lightweight finetuning. It is natural to wonder how well these models adapt to genuinely new tasks, but how does one find tasks that are unseen in internet-scale training sets? We turn to a field that is explicitly motivated and bottlenecked by a scarcity of web data: low-resource languages. In this paper, we introduce MTOB (Machine Translation from One Book), a benchmark for learning to translate between English and Kalamang—a language with less than 200 speakers and therefore virtually no presence on the web—using several hundred pages of field linguistics reference materials. This task framing is novel in that it asks a model to learn a language from a single human-readable book of grammar explanations, rather than a large mined corpus of in-domain data, more akin to L2 language learning than L1 language acquisition. We demonstrate that baselines using current LLMs are promising but fall short of human performance, achieving 44.7 chrF on Kalamang to English translation and 45.8 chrF on English to Kalamang translation, compared to 51.6 and 57.0 chrF by a human who learned Kalamang from the same reference materials. We hope that MTOB will help measure LLM capabilities along a new dimension, and that the methods developed to solve it could help expand access to language technology for underserved communities by leveraging qualitatively different kinds of data than traditional machine translation.

A Benchmark for Learning to Translate a New Language from One Grammar Book

2309.16575

https://openreview.net/forum?id=tbVWug9f2h

Poster

In recent years, data quality has emerged as an important factor for training massive models. Analytical theories suggest that higher-quality data can lead to lower test errors in models trained on a fixed data budget. Moreover, a model can be trained on a lower compute budget without compromising performance if a dataset can be stripped of its redundancies. Coreset selection (or data pruning) seeks to select a subset of the training data so as to maximize the performance of models trained on this subset, also referred to as coreset. There are two dominant approaches: (1) geometry-based data selection for maximizing *data diversity* in the coreset, and (2) functions that assign *difficulty scores* to samples based on training dynamics. Optimizing for data diversity leads to a coreset that is biased towards easier samples, whereas, selection by difficulty ranking omits easy samples that are necessary for the training of deep learning models. This demonstrates that data diversity and importance scores are two complementary factors that need to be jointly considered during coreset selection. In this work, we represent a dataset as an undirected graph and propose a novel pruning algorithm, $\mathbb{D}^2$ Pruning, that uses message passing over this dataset graph for coreset selection. $\mathbb{D}^2$ Pruning updates the difficulty scores of each example by incorporating the difficulty of its neighboring examples in the dataset graph. Then, these updated difficulty scores direct a graph-based sampling method to select a coreset that encapsulates both diverse and difficult regions of the dataset space. We evaluate supervised and self-supervised versions of our method on various vision and NLP datasets. Results show that $\mathbb{D}^2$ Pruning improves coreset selection over previous state-of-the-art methods at low-to-medium pruning rates. Additionally, we find that using $\mathbb{D}^2$ Pruning for filtering large multimodal datasets leads to increased diversity in the dataset and improved generalization of pretrained models. Our work shows that $\mathbb{D}^2$ Pruning is a versatile framework for understanding and processing datasets.

$\mathbb{D}^2$ Pruning: Message Passing for Balancing Diversity & Difficulty in Data Pruning

https://openreview.net/forum?id=thbtoAkCe9

Poster

As language models (LMs) become more capable, it is increasingly important to align them with human preferences. However, the dominant paradigm for training Preference Models (PMs) for that purpose suffers from fundamental limitations, such as lack of transparency and scalability, along with susceptibility to overfitting the preference dataset.We propose Compositional Preference Models (CPMs), a novel PM framework that decomposes one global preference assessment into several interpretable features, obtains scalar scores for these features from a prompted LM, and aggregates these scores using a logistic regression classifier. CPMs allow to control which properties of the preference data are used to train the preference model and to build it based on features that are believed to underlie the human preference judgment.Our experiments show that CPMs not only improve generalization and are more robust to overoptimization than standard PMs, but also that best-of-n samples obtained using CPMs tend to be preferred over samples obtained using conventional PMs.Overall, our approach demonstrates the benefits of endowing PMs with priors about which features determine human preferences while relying on LM capabilities to extract those features in a scalable and robust way.

Compositional Preference Models for Aligning LMs

2310.13011

https://openreview.net/forum?id=tiiAzqi6Ol

Spotlight Poster

Social reward as a form of community recognition provides a strong source ofmotivation for users of online platforms to actively engage and contribute withcontent to accumulate peers approval. In the realm of text-conditioned imagesynthesis, the recent surge in progress has ushered in a collaborative era whereusers and AI systems coalesce to refine visual creations. This co-creative pro-cess in the landscape of online social networks empowers users to craft originalvisual artworks seeking for community validation. Nevertheless, assessing thesemodels in the context of collective community preference introduces distinct chal-lenges. Existing evaluation methods predominantly center on limited size userstudies guided by image quality and alignment with prompts. This work pio-neers a paradigm shift, unveiling Social Reward - an innovative reward modelingframework that leverages implicit feedback from social network users engagedin creative editing of generated images. We embark on an extensive journey ofdataset curation and refinement, drawing from Picsart: an online visual creationand editing platform, yielding a first million-user-scale dataset of implicit humanpreferences for user-generated visual art named Picsart Image-Social. Our anal-ysis exposes the shortcomings of current metrics in modeling community creativepreference of text-to-image models’ outputs, compelling us to introduce a novelpredictive model explicitly tailored to address these limitations. Rigorous quan-titative experiments and user study show that our Social Reward model alignsbetter with social popularity than existing metrics. Furthermore, we utilize So-cial Reward to fine-tune text-to-image models, yielding images that are more fa-vored by not only Social Reward, but also other established metrics. These find-ings highlight the relevance and effectiveness of Social Reward in assessing com-munity appreciation for AI-generated artworks, establishing a closer alignmentwith users’ creative goals: creating popular visual art. Codes can be accessed athttps://github.com/Picsart-AI-Research/Social-Reward

Social Reward: Evaluating and Enhancing Generative AI through Million-User Feedback from an Online Creative Community

2402.09872

https://openreview.net/forum?id=tjn2YZSHUv

Poster

One-shot Federated Learning (OFL) has become a promising learning paradigm, enabling the training of a global server model via a single communication round. In OFL, the server model is aggregated by distilling knowledge from all client models (the ensemble), which are also responsible for synthesizing samples for distillation. In this regard, advanced works show that the performance of the server model is intrinsically related to the quality of the synthesized data and the ensemble model. To promote OFL, we introduce a novel framework, Co-Boosting, in which synthesized data and the ensemble model mutually enhance each other progressively. Specifically, Co-Boosting leverages the current ensemble model to synthesize higher-quality samples in an adversarial attack manner. These hard samples are then employed to promote the quality of the ensemble model by adjusting the ensembling weights for each client model. Consequently, Co-Boosting periodically achieves high-quality data and ensemble models. Extensive experiments demonstrate that Co-Boosting can substantially outperform existing baselines under various settings. Moreover, Co-Boosting eliminates the need for adjustments to the client's local training, requires no additional data or model transmission, and allows client models to have heterogeneous architectures.

Enhancing One-Shot Federated Learning Through Data and Ensemble Co-Boosting

2402.15070

https://openreview.net/forum?id=tm8s3696Ox

Poster

Exploring rich environments and evaluating one's actions without prior knowledge is immensely challenging. In this paper, we propose Motif, a general method to interface such prior knowledge from a Large Language Model (LLM) with an agent. Motif is based on the idea of grounding LLMs for decision-making without requiring them to interact with the environment: it elicits preferences from an LLM over pairs of captions to construct an intrinsic reward, which is then used to train agents with reinforcement learning. We evaluate Motif's performance and behavior on the challenging, open-ended and procedurally-generated NetHack game. Surprisingly, by only learning to maximize its intrinsic reward, Motif achieves a higher game score than an algorithm directly trained to maximize the score itself. When combining Motif's intrinsic reward with the environment reward, our method significantly outperforms existing approaches and makes progress on tasks where no advancements have ever been made without demonstrations. Finally, we show that Motif mostly generates intuitive human-aligned behaviors which can be steered easily through prompt modifications, while scaling well with the LLM size and the amount of information given in the prompt.

Motif: Intrinsic Motivation from Artificial Intelligence Feedback

2310.00166

https://openreview.net/forum?id=tmBKIecDE9

Spotlight Poster

*Multi-agent reinforcement learning* (MARL) has been shown effective for cooperative games in recent years. However, existing state-of-the-art methods face challenges related to sample complexity, training instability, and the risk of converging to a suboptimal Nash Equilibrium. In this paper, we propose a unified framework for learning \emph{stochastic} policies to resolve these issues. We embed cooperative MARL problems into probabilistic graphical models, from which we derive the maximum entropy (MaxEnt) objective for MARL. Based on the MaxEnt framework, we propose *Heterogeneous-Agent Soft Actor-Critic* (HASAC) algorithm. Theoretically, we prove the monotonic improvement and convergence to *quantal response equilibrium* (QRE) properties of HASAC. Furthermore, we generalize a unified template for MaxEnt algorithmic design named *Maximum Entropy Heterogeneous-Agent Mirror Learning* (MEHAML), which provides any induced method with the same guarantees as HASAC. We evaluate HASAC on six benchmarks: Bi-DexHands, Multi-Agent MuJoCo, StarCraft Multi-Agent Challenge, Google Research Football, Multi-Agent Particle Environment, and Light Aircraft Game. Results show that HASAC consistently outperforms strong baselines, exhibiting better sample efficiency, robustness, and sufficient exploration.

Maximum Entropy Heterogeneous-Agent Reinforcement Learning

2306.10715

https://openreview.net/forum?id=tmqOhBC4a5

Poster

Finetuning language models on domain-specific corpus is a common approach to enhance their domain knowledge and capability. While improving performance on domain tasks, it often brings a side-effect of forgetting of the model's general abilities. In this study, we analyze the effects of finetuning on language models by dissecting its impacts on the modeling of topic, style, and factual knowledge in text. Our method uses instruction-following LLMs such as ChatGPT to auto-generate controlled-variable text examples which we use to probe the model. Our findings reveal that finetuning results in significant shifts in the language model's topic and style priors, while actual knowledge learning only contributes to a small fraction of the total probability change. Analysis shows that the adaptation of topic and style priors behave akin to learning simple features: they are learned rapidly and require little model capacity. They are also learned independently and primarily at the beginning of a text sequence. In contrast, factual knowledge is learned stably but slowly and requires significant model capacity to learn. The research offers insights and understanding into the finer dynamics of learning and forgetting in language models, and can potentially inform future research on improving domain adaptation and addressing the challenges of forgetting in continual learning of language models.

Dissecting learning and forgetting in language model finetuning

https://openreview.net/forum?id=tmsqb6WpLz

Poster

Foundation models like CLIP are trained on hundreds of millions of samples and effortlessly generalize to new tasks and inputs. Out of the box, CLIP shows stellar zero-shot and few-shot capabilities on a wide range of out-of-distribution (OOD) benchmarks, which prior works attribute mainly to today's large and comprehensive training dataset (like LAION). However, it is questionable how meaningful terms like out-of-distribution generalization are for CLIP as it seems likely that web-scale datasets like LAION simply contain many samples that are similar to common OOD benchmarks originally designed for ImageNet. To test this hypothesis, we retrain CLIP on pruned LAION splits that replicate ImageNet’s train-test similarity with respect to common OOD benchmarks. While we observe a performance drop on some benchmarks, surprisingly, CLIP’s overall performance remains high. This shows that high train-test similarity is insufficient to explain CLIP’s OOD performance, and other properties of the training data must drive CLIP to learn more generalizable representations. Additionally, by pruning data points that are dissimilar to the OOD benchmarks, we uncover a 100M split of LAION (¼ of its original size) on which CLIP can be trained to match its original OOD performance.

Does CLIP’s generalization performance mainly stem from high train-test similarity?

https://openreview.net/forum?id=tnBaiidobu

Poster

Diffusion models have achieved tremendous success in generating high-dimensional data like images, videos and audio. These models provide powerful data priors that can solve linear inverse problems in zero shot through Bayesian posterior sampling.However, exact posterior sampling for diffusion models is intractable. Current solutions often hinge on approximations that are either computationally expensive or lack strong theoretical guarantees. In this work, we introduce an efficient diffusion sampling algorithm for linear inverse problems that is guaranteed to be asymptotically accurate. We reveal a link between Bayesian posterior sampling and Bayesian filtering in diffusion models, proving the former as a specific instance of the latter. Our method, termed filtering posterior sampling, leverages sequential Monte Carlo methods to solve the corresponding filtering problem. It seamlessly integrates with all Markovian diffusion samplers, requires no model re-training, and guarantees accurate samples from the Bayesian posterior as particle counts rise. Empirical tests demonstrate that our method generates better or comparable results than leading zero-shot diffusion posterior samplers on tasks like image inpainting, super-resolution, and deblurring.

Diffusion Posterior Sampling for Linear Inverse Problem Solving: A Filtering Perspective

https://openreview.net/forum?id=tplXNcHZs1

Poster

As research in large language models (LLMs) continues to accelerate, LLM-based evaluation has emerged as a scalable and cost-effective alternative to human evaluations for comparing the ever-increasing list of models. This paper investigates the efficacy of these “LLM evaluators”, particularly in using them to assess instruction following, a metric that gauges how closely generated text adheres to theinstructions. We introduce a challenging meta-evaluation benchmark, LLMBAR, designed to test the ability of an LLM evaluator to discern instruction-following outputs. The authors curated 419 pairs of outputs, one adhering to instructions while the other diverging, yet may possess deceptive qualities that could mislead an LLM evaluator. Contrary to existing meta-evaluation, we discover that different evaluators (i.e., combinations of LLMs and prompts) exhibit distinct performance on LLMBAR and even the highest-scoring LLM evaluators have substantial room for improvement. We also present a novel suite of prompting strategies that further close the gap between LLM and human evaluators. With LLMBAR, we hope to offer more insight into the behavior of LLM evaluators and foster research in developing better instruction-following models.

Evaluating Large Language Models at Evaluating Instruction Following

2310.07641

https://openreview.net/forum?id=tr0KidwPLc

Poster

The deployment of Reinforcement Learning (RL) in real-world applications is constrained by its failure to satisfy safety criteria.Existing Safe Reinforcement Learning (SafeRL) methods, which rely on cost functions to enforce safety, often fail to achieve zero-cost performance in complex scenarios, especially vision-only tasks. These limitations are primarily due to model inaccuracies and inadequate sample efficiency. The integration of world models has proven effective in mitigating these shortcomings. In this work, we introduce SafeDreamer, a novel algorithm incorporating Lagrangian-based methods into world model planning processes within the superior Dreamer framework. Our method achieves nearly zero-cost performance on various tasks, spanning low-dimensional and vision-only input, within the Safety-Gymnasium benchmark, showcasing its efficacy in balancing performance and safety in RL tasks. Further details and resources are available on the project website: https://sites.google.com/view/safedreamer.

SafeDreamer: Safe Reinforcement Learning with World Models

2307.07176

https://openreview.net/forum?id=tsE5HLYtYg

Poster

Building cross-modal applications is challenging due to limited paired multi-modal data. Recent works have shown that leveraging a pre-trained multi-modal contrastive representation space enables cross-modal tasks to be learned from uni-modal data. This is based on the assumption that contrastive optimization makes embeddings from different modalities interchangeable. However, this assumption is under-explored due to the poorly understood geometry of the multi-modal contrastive space, where a modality gap exists. In our study, we provide a theoretical explanation of this space's geometry and introduce a three-step method, $C^3$ (Connect, Collapse, Corrupt), to bridge the modality gap, enhancing the interchangeability of embeddings. Our $C^3$ method significantly improves cross-modal learning from uni-modal data, achieving state-of-the-art results on zero-shot image / audio / video captioning and text-to-image generation.

Connect, Collapse, Corrupt: Learning Cross-Modal Tasks with Uni-Modal Data

2401.08567

https://openreview.net/forum?id=ttXg3SKAg5

Poster

Pipeline parallelism is one of the key components for large-scale distributed training, yet its efficiency suffers from pipeline bubbles which were deemed inevitable. In this work, we introduce a scheduling strategy that, to our knowledge, is the first to successfully achieve zero pipeline bubbles under synchronous training semantics. The key idea behind this improvement is to split the backward computation into two parts, one that computes gradient for the input and another that computes for the parameters. Based on this idea, we handcraft novel pipeline schedules that significantly outperform the baseline methods. We further develop an algorithm that automatically finds an optimal schedule based on specific model configuration and memory limit. Additionally, to truly achieve zero bubble, we introduce a novel technique to bypass synchronizations during the optimizer step. Experimental evaluations show that our method outperforms the 1F1B schedule up to 15\% in throughput under a similar memory limit. This number can be further pushed to 30\% when the memory constraint is relaxed. We believe our results mark a major step forward in harnessing the true potential of pipeline parallelism. The source code based on Megatron-LM is publicly avaiable at \url{https://github.com/sail-sg/zero-bubble-pipeline-parallelism}.

Zero Bubble (Almost) Pipeline Parallelism

https://openreview.net/forum?id=tuzTN0eIO5

Spotlight Poster

Training-free metrics (a.k.a. zero-cost proxies) are widely used to avoid resource-intensive neural network training, especially in Neural Architecture Search (NAS). Recent studies show that existing training-free metrics have several limitations, such as limited correlation and poor generalisation across different search spaces and tasks. Hence, we propose a novel high-performance training-free metric, SWAP-Score, based on Sample-Wise Activation Patterns. It measures the expressivity of networks over a batch of input samples. The SWAP-Score is strongly correlated with ground-truth performance across various search spaces and tasks, outperforming 15 existing training-free metrics on NAS-Bench-101/201/301 and TransNAS-Bench-101. The SWAP-Score can be further enhanced by regularisation, which leads to even higher correlations in cell-based search space and enables model size control during the search. For example, Spearman's rank correlation coefficient between regularised SWAP-Score and CIFAR-100 validation accuracies on NAS-Bench-201 networks is 0.90, significantly higher than 0.80 from the second-best metric, NWOT. When integrated with an evolutionary algorithm for NAS, our SWAP-NAS achieves competitive performance on CIFAR-10 and ImageNet in approximately 6 minutes and 9 minutes of GPU time respectively.

SWAP-NAS: Sample-Wise Activation Patterns for Ultra-fast NAS

https://openreview.net/forum?id=tveiUXU2aa

Poster

Reinforcement learning from human feedback (RLHF) is a popular technique for training high-quality AI assistants. However, RLHF may also encourage model responses that match user beliefs over truthful responses, a behavior known as sycophancy. We investigate the prevalence of sycophancy in RLHF-trained models and whether human preference judgments are responsible. We first demonstrate that five state-of-the-art AI assistants consistently exhibit sycophancy behavior across four varied free-form text-generation tasks. To understand if human preferences drive this broadly observed behavior of RLHF models, we analyze existing human preference data. We find that when a response matches a user's views, it is more likely to be preferred. Moreover, both humans and preference models (PMs) prefer convincingly-written sycophantic responses over correct ones a non-negligible fraction of the time. Optimizing model outputs against PMs also sometimes sacrifices truthfulness in favor of sycophancy. Overall, our results indicate that sycophancy is a general behavior of RLHF models, likely driven in part by human preference judgments favoring sycophantic responses.

Towards Understanding Sycophancy in Language Models

2310.13548

https://openreview.net/forum?id=tvhaxkMKAn

Poster

We consider the general class of time-homogeneous stochastic dynamical systems,both discrete and continuous, and study the problem of learning a representationof the state that faithfully captures its dynamics. This is instrumental to learn thetransfer operator of the system, that in turn can be used for numerous tasks, suchas forecasting and interpreting the system dynamics. We show that the searchfor a good representation can be cast as an optimization problem over neuralnetworks. Our approach is supported by recent results in statistical learning theory,highlighting the role of approximation error and metric distortion in the context oftransfer operator regression. The objective function we propose is associated withprojection operators from the representation space to the data space, overcomesmetric distortion, and can be empirically estimated from data. In the discrete timesetting, we further derive a relaxed objective function that is differentiable andnumerically well-conditioned. We compare our method against state-of-the-artapproaches on different datasets, showing better performance across the board

Learning invariant representations of time-homogeneous stochastic dynamical systems

2307.09912

https://openreview.net/forum?id=twSnZwiOIm

Poster

In continual learning, catastrophic forgetting is affected by multiple aspects of the tasks. Previous works have analyzed separately how forgetting is affected by either task similarity or overparameterization. In contrast, our paper examines how task similarity and overparameterization jointly affect forgetting in an analyzable model. Specifically, we focus on two-task continual linear regression, where the second task is a random orthogonal transformation of an arbitrary first task (an abstraction of random permutation tasks). We derive an exact analytical expression for the expected forgetting — and uncover a nuanced pattern. In highly overparameterized models, intermediate task similarity causes the most forgetting. However, near the interpolation threshold, forgetting decreases monotonically with the expected task similarity. We validate our findings with linear regression on synthetic data, and with neural networks on established permutation task benchmarks.

The Joint Effect of Task Similarity and Overparameterization on Catastrophic Forgetting — An Analytical Model

2401.12617

https://openreview.net/forum?id=u3dHl287oB

Poster

Imposing known physical constraints, such as conservation laws, during neural network training introduces an inductive bias that can improve accuracy, reliability, convergence, and data efficiency for modeling physical dynamics. While such constraints can be softly imposed via loss function penalties, recent advancements in differentiable physics and optimization improve performance by incorporating PDE-constrained optimization as individual layers in neural networks. This enables a stricter adherence to physical constraints. However, imposing hard constraints significantly increases computational and memory costs, especially for complex dynamical systems. This is because it requires solving an optimization problem over a large number of points in a mesh, representing spatial and temporal discretizations, which greatly increases the complexity of the constraint. To address this challenge, we develop a scalable approach to enforce hard physical constraints using Mixture-of-Experts (MoE), which can be used with any neural network architecture. Our approach imposes the constraint over smaller decomposed domains, each of which is solved by an ``expert'' through differentiable optimization. During training, each expert independently performs a localized backpropagation step by leveraging the implicit function theorem; the independence of each expert allows for parallelization across multiple GPUs. Compared to standard differentiable optimization, our scalable approach achieves greater accuracy in the neural PDE solver setting for predicting the dynamics of challenging non-linear systems. We also improve training stability and require significantly less computation time during both training and inference stages.

Scaling physics-informed hard constraints with mixture-of-experts

2402.13412

https://openreview.net/forum?id=u3dX2CEIZb

Spotlight Poster

In this work, we investigate the capability of generating images from pre-trained diffusion models at much higher resolutions than the training image sizes. In addition, the generated images should have arbitrary image aspect ratios. When generating images directly at a higher resolution, 1024 x 1024, with the pre-trained Stable Diffusion using training images of resolution 512 x 512, we observe persistent problems of object repetition and unreasonable object structures. Existing works for higher-resolution generation, such as attention-based and joint-diffusion approaches, cannot well address these issues. As a new perspective, we examine the structural components of the U-Net in diffusion models and identify the crucial cause as the limited perception field of convolutional kernels. Based on this key observation, we propose a simple yet effective re-dilation that can dynamically adjust the convolutional perception field during inference. We further propose the dispersed convolution and noise-damped classifier-free guidance, which can enable ultra-high-resolution image generation (e.g., 4096 x 4096). Notably, our approach does not require any training or optimization. Extensive experiments demonstrate that our approach can address the repetition issue well and achieve state-of-the-art performance on higher-resolution image synthesis, especially in texture details. Our work also suggests that a pre-trained diffusion model trained on low-resolution images can be directly used for high-resolution visual generation without further tuning, which may provide insights for future research on ultra-high-resolution image and video synthesis. More results are available at the anonymous website: https://scalecrafter.github.io/ScaleCrafter/

ScaleCrafter: Tuning-free Higher-Resolution Visual Generation with Diffusion Models

2310.07702

https://openreview.net/forum?id=u48tHG5f66

Poster

We show that large language models (LLMs) can be adapted to be generalizable policies for embodied visual tasks. Our approach, called Large LAnguage model Reinforcement Learning Policy (LLaRP), adapts a pre-trained frozen LLM to take as input text instructions and visual egocentric observations and output actions directly in the environment. Using reinforcement learning, we train LLaRP to see and act solely through environmental interactions. We show that LLaRP is robust to complex paraphrasings of task instructions and can generalize to new tasks that require novel optimal behavior. In particular, on 1,000 unseen tasks it achieves 42% success rate, 1.7x the success rate of other common learned baselines or zero-shot applications of LLMs. Finally, to aid the community in studying language conditioned, massively multi-task, embodied AI problems we release a novel benchmark, Language Rearrangement, consisting of 150,000 training and 1,000 testing tasks for language-conditioned rearrangement.

Large Language Models as Generalizable Policies for Embodied Tasks

2310.17722

https://openreview.net/forum?id=u6imHU4Ebu

Poster

The deep neural networks are known to be vulnerable to well-designed adversarial attacks. The most successful defense technique based on adversarial training (AT) can achieve optimal robustness against particular attacks but cannot generalize well to unseen attacks. Another effective defense technique based on adversarial purification (AP) can enhance generalization but cannot achieve optimal robustness. Meanwhile, both methods share one common limitation on the degraded standard accuracy. To mitigate these issues, we propose a novel framework called Adversarial Training on Purification (AToP), which comprises two components: perturbation destruction by random transforms (RT) and purifier model fine-tuned (FT) by adversarial loss. RT is essential to avoid overlearning to known attacks resulting in the robustness generalization to unseen attacks and FT is essential for the improvement of robustness. To evaluate our method in an efficient and scalable way, we conduct extensive experiments on CIFAR-10, CIFAR-100, and ImageNette to demonstrate that our method achieves state-of-the-art results and exhibits generalization ability against unseen attacks.

Adversarial Training on Purification (AToP): Advancing Both Robustness and Generalization

2401.16352

https://openreview.net/forum?id=u7559ZMvwY

Spotlight Poster

Manipulating objects is a hallmark of human intelligence, and an important task in domains such as robotics. In principle, Reinforcement Learning (RL) offers a general approach to learn object manipulation. In practice, however, domains with more than a few objects are difficult for RL agents due to the curse of dimensionality, especially when learning from raw image observations. In this work we propose a structured approach for visual RL that is suitable for representing multiple objects and their interaction, and use it to learn goal-conditioned manipulation of several objects. Key to our method is the ability to handle goals with dependencies between the objects (e.g., moving objects in a certain order). We further relate our architecture to the generalization capability of the trained agent, and demonstrate agents that learn with 3 objects but generalize to similar tasks with over 10 objects. Rollout videos are available on our website: https://sites.google.com/view/entity-centric-rl

Entity-Centric Reinforcement Learning for Object Manipulation from Pixels

2404.01220

https://openreview.net/forum?id=uDxeSZ1wdI

Poster

We present ReCAT, a recursive composition augmented Transformer that is able to explicitly model hierarchical syntactic structures of raw texts without relying on gold trees during both learning and inference. Existing research along this line restricts data to follow a hierarchical tree structure and thus lacks inter-span communications.To overcome the problem, we propose a novel contextual inside-outside (CIO) layer that learns contextualized representations of spans through bottom-up and top-down passes, where a bottom-up pass forms representations of high-level spans by composing low-level spans, while a top-down pass combines information inside and outside a span. By stacking several CIO layers between the embedding layer and the attention layers in Transformer, the ReCAT model can perform both deep intra-span and deep inter-span interactions, and thus generate multi-grained representations fully contextualized with other spans.Moreover, the CIO layers can be jointly pre-trained with Transformers, making ReCAT enjoy scaling ability, strong performance, and interpretability at the same time. We conduct experiments on various sentence-level and span-level tasks. Evaluation results indicate that ReCAT can significantly outperform vanilla Transformer models on all span-level tasks and recursive models on natural language inference tasks. More interestingly, the hierarchical structures induced by ReCAT exhibit strong consistency with human-annotated syntactic trees, indicating good interpretability brought by the CIO layers.

Augmenting Transformers with Recursively Composed Multi-grained Representations

2309.16319

https://openreview.net/forum?id=u859gX7ADC

Spotlight Poster

The current face recognition (FR) algorithms has achieved a high level of accuracy, making further improvements increasingly challenging. While existing FR algorithms primarily focus on optimizing margins and loss functions, limited attention has been given to exploring the feature representation space. Therefore, this paper endeavors to improve FR performance in the view of feature representation space. Firstly, we consider two FR models that exhibit distinct performance discrepancies, where one model exhibits superior recognition accuracy compared to the other. We implement orthogonal decomposition on the features from the superior model along those from the inferior model and obtain two sub-features. Surprisingly, we find the sub-feature perpendicular to the inferior still possesses a certain level of face distinguishability. We adjust the modulus of the sub-features and recombine them through vector addition. Experiments demonstrate this recombination is likely to contribute to an improved facial feature representation, even better than features from the original superior model. Motivated by this discovery, we further consider how to improve FR accuracy when there is only one FR model available. Inspired by knowledge distillation, we incorporate the intra-class incoherence constraint (IIC) to solve the problem. Experiments on various FR benchmarks show the existing state-of-the-art method with IIC can be further improved, highlighting its potential to further enhance FR performance.

Enhanced Face Recognition using Intra-class Incoherence Constraint

https://openreview.net/forum?id=uELjxVbrqG

Poster

Differentially Private Stochastic Gradient Descent with gradient clipping (DPSGD-GC) is a powerful tool for training deep learning models using sensitive data, providing both a solid theoretical privacy guarantee and high efficiency. However, existing research has shown that DPSGD-GC only converges when using large clipping thresholds that are dependent on problem-specific parameters that are often unknown in practice. Therefore, DPSGD-GC suffers from degraded performance due to the {\it constant} bias introduced by the clipping. In our work, we propose a new error-feedback (EF) DP algorithm as an alternative to DPSGD-GC, which offers a diminishing utility bound without inducing a constant clipping bias. More importantly, it allows for an arbitrary choice of clipping threshold that is independent of the problem. We establish an algorithm-specific DP analysis for our proposed algorithm, providing privacy guarantees based on R{\'e}nyi DP. And we demonstrate that under mild conditions, our algorithm can achieve nearly the same utility bound as DPSGD without gradient clipping. Our empirical results on standard datasets show that the proposed algorithm achieves higher accuracies than DPSGD while maintaining the same level of DP guarantee.

Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach

2311.14632

https://openreview.net/forum?id=uFbWHyTlPn

Poster

The remarkable success in neural networks provokes the selective rationalization. It explains the prediction results by identifying a small subset of the inputs sufficient to support them. Since existing methods still suffer from adopting the shortcuts in data to compose rationales and limited large-scale annotated rationales by human, in this paper, we propose a Shortcuts-fused Selective Rationalization (SSR) method, which boosts the rationalization by discovering and exploiting potential shortcuts. Specifically, SSR first designs a shortcuts discovery approach to detect several potential shortcuts. Then, by introducing the identified shortcuts, we propose two strategies to mitigate the problem of utilizing shortcuts to compose rationales. Finally, we develop two data augmentations methods to close the gap in the number of annotated rationales. Extensive experimental results on four real-world datasets clearly validate the effectiveness of our proposed method.

Towards Faithful Explanations: Boosting Rationalization with Shortcuts Discovery

2403.07955

https://openreview.net/forum?id=uGtfk2OphU

Poster

Generative flow networks (GFlowNets) are sequential sampling models trained to match a given distribution. GFlowNets have been successfully applied to various structured object generation tasks, sampling a diverse set of high-reward objects quickly. We propose expected flow networks (EFlowNets), which extend GFlowNets to stochastic environments. We show that EFlowNets outperform other GFlowNet formulations in stochastic tasks such as protein design. We then extend the concept of EFlowNets to adversarial environments, proposing adversarial flow networks (AFlowNets) for two-player zero-sum games. We show that AFlowNets learn to find above 80% of optimal moves in Connect-4 via self-play and outperform AlphaZero in tournaments.

Expected flow networks in stochastic environments and two-player zero-sum games

2310.02779

https://openreview.net/forum?id=uH0FGECSEI

Poster

In-context learning is a promising approach for online policy learning of offline reinforcement learning (RL) methods, which can be achieved at inference time without gradient optimization. However, this method is hindered by significant computational costs resulting from the gathering of large training trajectory sets and the need to train large Transformer models. We address this challenge by introducing an In-context Exploration-Exploitation (ICEE) algorithm, designed to optimize the efficiency of in-context policy learning. Unlike existing models, ICEE performs an exploration-exploitation trade-off at inference time within a Transformer model, without the need for explicit Bayesian inference. Consequently, ICEE can solve Bayesian optimization problems as efficiently as Gaussian process biased methods do, but in significantly less time. Through experiments in grid world environments, we demonstrate that ICEE can learn to solve new RL tasks using only tens of episodes, marking a substantial improvement over the hundreds of episodes needed by the previous in-context learning method.

In-context Exploration-Exploitation for Reinforcement Learning

2403.06826

https://openreview.net/forum?id=uIKZSStON3

Poster

As transformers are equivariant to the permutation of input tokens, encoding the positional information of tokens is necessary for many tasks. However, since existing positional encoding schemes have been initially designed for NLP tasks, their suitability for vision tasks, which typically exhibit different structural properties in their data, is questionable. We argue that existing positional encoding schemes are suboptimal for 3D vision tasks, as they do not respect their underlying 3D geometric structure. Based on this hypothesis, we propose a geometry-aware attention mechanism that encodes the geometric structure of tokens as relative transformation determined by the geometric relationship between queries and key-value pairs. By evaluating on multiple novel view synthesis (NVS) datasets in the sparse wide-baseline multi-view setting, we show that our attention, called Geometric Transform Attention (GTA), improves learning efficiency and performance of state-of-the-art transformer-based NVS models without any additional learned parameters and only minor computational overhead.

GTA: A Geometry-Aware Attention Mechanism for Multi-View Transformers

2310.10375

https://openreview.net/forum?id=uJVHygNeSZ

Poster

The genome sequence contains the blueprint for governing cellular processes. While the availability of genomes has vastly increased over the last decades, experimental annotation of the various functional, non-coding and regulatory elements encoded in the DNA sequence remains both expensive and challenging. This has sparked interest in unsupervised language modeling of genomic DNA, a paradigm that has seen great success for protein sequence data. Although various DNA language models have been proposed, evaluation tasks often differ between individual works, and might not fully recapitulate the fundamental challenges of genome annotation, including the length, scale and sparsity of the data. In this study, we introduce **BEND**, a **BEN**chmark for **D**NA language models, featuring a collection of realistic and biologically meaningful downstream tasks defined on the human genome. We find that embeddings from current DNA LMs can approach performance of expert methods on some tasks, but only capture limited information about long-range features. BEND is available at https://anonymous.4open.science/r/BEND-8C42/README.md

BEND: Benchmarking DNA Language Models on Biologically Meaningful Tasks

2311.12570

https://openreview.net/forum?id=uKB4cFNQFg

Poster

Pocket representations play a vital role in various biomedical applications, such as druggability estimation, ligand affinity prediction, and de novo drug design. While existing geometric features and pretrained representations have demonstrated promising results, they usually treat pockets independent of ligands, neglecting the fundamental interactions between them. However, the limited pocket-ligand complex structures available in the PDB database (less than 100 thousand non-redundant pairs) hampers large-scale pretraining endeavors for interaction modeling. To address this constraint, we propose a novel pocket pretraining approach that leverages knowledge from high-resolution atomic protein structures, assisted by highly effective pretrained small molecule representations. By segmenting protein structures into drug-like fragments and their corresponding pockets, we obtain a reasonable simulation of ligand-receptor interactions, resulting in the generation of over 5 million complexes. Subsequently, the pocket encoder is trained in a contrastive manner to align with the representation of pseudo-ligand furnished by some pretrained small molecule encoders. Our method, named ProFSA, achieves state-of-the-art performance across various tasks, including pocket druggability prediction, pocket matching, and ligand binding affinity prediction. Notably, ProFSA surpasses other pretraining methods by a substantial margin. Moreover, our work opens up a new avenue for mitigating the scarcity of protein-ligand complex data through the utilization of high-quality and diverse protein structure databases.

Self-supervised Pocket Pretraining via Protein Fragment-Surroundings Alignment

2310.07229

https://openreview.net/forum?id=uMAujpVi9m

Poster

Out-of-distribution (OOD) detection aims to detect testing samples far away from the in-distribution (ID) training data, which is crucial for the safe deployment of machine learning models in the real world. Distance-based OOD detection methods have emerged with enhanced deep representation learning. They identify unseen OOD samples by measuring their distances from ID class centroids or prototypes. However, existing approaches learn the representation relying on oversimplified data assumptions, e.g. modeling ID data of each class with one centroid class prototype or using loss functions not designed for OOD detection, which overlook the natural diversities within the data. Naively enforcing data samples of each class to be compact around only one prototype leads to inadequate modeling of realistic data and limited performance. To tackle these issues, we propose PrototypicAl Learning with a Mixture of prototypes (PALM) that models each class with multiple prototypes to capture the sample diversities, which learns more faithful and compact samples embeddings for enhanching OOD detection. Our method automatically identifies and dynamically updates prototypes, assigning each sample to a subset of prototypes via reciprocal neighbor soft assignment weights. To learn embeddings with multiple prototypes, PALM optimizes a maximum likelihood estimation (MLE) loss to encourage the sample embeddings to compact around the associated prototypes, as well as a contrastive loss on all prototypes to enhance intra-class compactness and inter-class discrimination at the prototype level. Compared to previous methods with prototypes, the proposed mixture prototype modeling of PALM promotes the representations of each ID class to be more compact and separable from others and the unseen OOD samples, resulting in more reliable OOD detection. Moreover, the automatic estimation of prototypes enables our approach to be extended to the challenging OOD detection task with unlabelled ID data. Extensive experiments demonstrate the superiority of PALM over previous methods, achieving state-of-the-art average AUROC performance of 93.82 on the challenging CIFAR-100 benchmark.

Learning with Mixture of Prototypes for Out-of-Distribution Detection

2402.02653

https://openreview.net/forum?id=uNkKaD3MCs

Poster

Diffusion models are a powerful class of generative models which simulate stochastic differential equations (SDEs) to generate data from noise. Although diffusion models have achieved remarkable progress in recent years, they have limitations in the unpaired image-to-image translation tasks due to the Gaussian prior assumption. Schrödinger Bridge (SB), which learns an SDE to translate between two arbitrary distributions, have risen as an attractive solution to this problem. However, none of SB models so far have been successful at unpaired translation between high-resolution images. In this work, we propose the Unpaired Neural Schrödinger Bridge (UNSB), which expresses SB problem as a sequence of adversarial learning problems. This allows us to incorporate advanced discriminators and regularization to learn a SB between unpaired data. We demonstrate that UNSB is scalable and successfully solves various unpaired image-to-image translation tasks.

Unpaired Image-to-Image Translation via Neural Schrödinger Bridge

https://openreview.net/forum?id=uQBW7ELXfO

Poster

We propose the In-context Autoencoder (ICAE), leveraging the power of a large language models (LLM) to compress a long context into short compact memory slots that can be directly conditioned on by the LLM for various purposes. ICAE is first pretrained using both autoencoding and language modeling objectives on massive text data, enabling it to generate memory slots that accurately and comprehensively represent the original context; Then, it is fine-tuned on instruction data for producing desirable responses to various prompts. Experiments demonstrate that our lightweight ICAE, introducing fewer than 1% additional parameters, effectively achieves $4\times$ context compression based on Llama, offering advantages in both improved latency and GPU memory cost during inference, and showing an interesting insight in memorization as well as potential for scalability. These promising results imply a novel perspective on the connection between working memory in cognitive science and representation learning in LLMs, revealing ICAE's significant implications in addressing the long context problem and suggesting further research in LLM context management. Our data, code and model will be released.

In-context Autoencoder for Context Compression in a Large Language Model

2307.06945

https://openreview.net/forum?id=uREj4ZuGJE

Spotlight Poster

The recent advancements in large language models (LLMs) have sparked a growing apprehension regarding the potential misuse. One approach to mitigating this risk is to incorporate watermarking techniques into LLMs, allowing for the tracking and attribution of model outputs. This study examines a crucial aspect of watermarking: how significantly watermarks impact the quality of model-generated outputs. Previous studies have suggested a trade-off between watermark strength and output quality. However, our research demonstrates that it is possible to integrate watermarks without affecting the output probability distribution with appropriate implementation. We refer to this type of watermark as an unbiased watermark. This has significant implications for the use of LLMs, as it becomes impossible for users to discern whether a service provider has incorporated watermarks or not. Furthermore, the presence of watermarks does not compromise the performance of the model in downstream tasks, ensuring that the overall utility of the language model is preserved. Our findings contribute to the ongoing discussion around responsible AI development, suggesting that unbiased watermarks can serve as an effective means of tracking and attributing model outputs without sacrificing output quality.

Unbiased Watermark for Large Language Models

2310.10669

https://openreview.net/forum?id=uWVC5FVidc

Poster

LoRA has gained widespread acceptance in the fine-tuning of large pre-trained models to cater to a diverse array of downstream tasks, showcasing notable effectiveness and efficiency, thereby solidifying its position as one of the most prevalent fine-tuning techniques. Due to the modular nature of LoRA's plug-and-play plugins, researchers have delved into the amalgamation of multiple LoRAs to empower models to excel across various downstream tasks. Nonetheless, extant approaches for LoRA fusion grapple with inherent challenges. Direct arithmetic merging may result in the loss of the original pre-trained model's generative capabilities or the distinct identity of LoRAs, thereby yielding suboptimal outcomes. On the other hand, Reference tuning-based fusion exhibits limitations concerning the requisite flexibility for the effective combination of multiple LoRAs. In response to these challenges, this paper introduces the Mixture of LoRA Experts (MoLE) approach, which harnesses hierarchical control and unfettered branch selection. The MoLE approach not only achieves superior LoRA fusion performance in comparison to direct arithmetic merging but also retains the crucial flexibility for combining LoRAs effectively. Extensive experimental evaluations conducted in both the Natural Language Processing (NLP) and Vision \& Language (V\&L) domains substantiate the efficacy of MoLE.

MoLE: Mixture of LoRA Experts

https://openreview.net/forum?id=uWvKBCYh4S

Poster

The vulnerability of deep neural networks to adversarial patches has motivated numerous defense strategies for boosting model robustness. However, the prevailing defenses depend on single observation or pre-established adversary information to counter adversarial patches, often failing to be confronted with unseen or adaptive adversarial attacks and easily exhibiting unsatisfying performance in dynamic 3D environments. Inspired by active human perception and recurrent feedback mechanisms, we develop Embodied Active Defense (EAD), a proactive defensive strategy that actively contextualizes environmental information to address misaligned adversarial patches in 3D real-world settings. To achieve this, EAD develops two central recurrent sub-modules, i.e., a perception module and a policy module, to implement two critical functions of active vision. These models recurrently process a series of beliefs and observations, facilitating progressive refinement of their comprehension of the target object and enabling the development of strategic actions to counter adversarial patches in 3D environments. To optimize learning efficiency, we incorporate a differentiable approximation of environmental dynamics and deploy patches that are agnostic to the adversary’s strategies. Extensive experiments demonstrate that EAD substantially enhances robustness against a variety of patches within just a few steps through its action policy in safety-critical tasks (e.g., face recognition and object detection), without compromising standard accuracy. Furthermore, due to the attack-agnostic characteristic, EAD facilitates excellent generalization to unseen attacks, diminishing the averaged attack success rate by $95\%$ across a range of unseen adversarial attacks.

Embodied Active Defense: Leveraging Recurrent Feedback to Counter Adversarial Patches

2404.00540

https://openreview.net/forum?id=uXjfOmTiDt

Poster

Implicit neural representations (INRs) have arisen as useful methods for representing signals on Euclidean domains. By parameterizing an image as a multilayer perceptron (MLP) on Euclidean space, INRs effectively represent signals in a way that couples spatial and spectral features of the signal that is not obvious in the usual discrete representation, paving the way for continuous signal processing and machine learning approaches that were not previously possible. Although INRs using sinusoidal activation functions have been studied in terms of Fourier theory, recent works have shown the advantage of using wavelets instead of sinusoids as activation functions, due to their ability to simultaneously localize in both frequency and space. In this work, we approach such INRs and demonstrate how they resolve high-frequency features of signals from coarse approximations done in the first layer of the MLP. This leads to multiple prescriptions for the design of INR architectures, including the use of complex wavelets, decoupling of low and band-pass approximations, and initialization schemes based on the singularities of the desired signal.

Implicit Neural Representations and the Algebra of Complex Wavelets

2310.00545

https://openreview.net/forum?id=uZfjFyPAvn

Poster

In deep metric learning, the triplet loss has emerged as a popular method to learn many computer vision and natural language processing tasks such as facial recognition, object detection, and visual-semantic embeddings. One issue that plagues the triplet loss is network collapse, an undesirable phenomenon where the network projects the embeddings of all data onto a single point. Researchers predominately solve this problem by using triplet mining strategies. While hard negative mining is the most effective of these strategies, existing formulations lack strong theoretical justification for their empirical success. In this paper, we utilize the mathematical theory of isometric approximation to show an equivalence between the triplet loss sampled by hard negative mining and an optimization problem that minimizes a Hausdorff-like distance between the neural network and its ideal counterpart function. This provides the theoretical justifications for hard negative mining's empirical efficacy. Experiments performed on the Market-1501 and Stanford Online Products datasets with various network architectures corroborate our theoretical findings, indicating that network collapse tends to happen when batch size is too large or embedding dimension is too small. In addition, our novel application of the isometric approximation theorem provides the groundwork for future forms of hard negative mining that avoid network collapse.

Mathematical Justification of Hard Negative Mining via Isometric Approximation Theorem

2210.11173

https://openreview.net/forum?id=udO3k28bEw

Poster

Large language models (LLMs) have established new standards in various natural language processing tasks. However, a primary constraint they face is the context limit, i.e., the maximum number of tokens they can process.To relax the constraint, previous works have explored architectural changes and modifications in positional encoding, but they often require expensive training or do not address the computational demands of self-attention.In this paper, we present Hierarchical cOntext MERging (HOMER), a new training-freescheme designed to overcome the limitations. HOMER harnesses a divide-and-conquer methodology, segmenting extensive inputs into manageable units. The segments are then processed collectively, employing a hierarchical strategy that fuses adjacent chunks at progressive Transformer layers. A token reduction technique precedes each fusion, ensuring memory usage efficiency.We also propose an optimized computational order reducing the memory requirement to logarithmically scale with respect to input length, making it especially favorable for environments with tight memory restrictions. Our experimental results demonstrate the superior performance and memory efficiency of the proposed method, opening doors for broader applications of LLMs in scenarios with extended context requirements.

Hierarchical Context Merging: Better Long Context Understanding for Pre-trained LLMs

2404.10308

https://openreview.net/forum?id=ulaUJFd96G

Spotlight Poster

We study the problem of building a controller that can follow open-ended instructions in open-world environments. We propose to follow reference videos as instructions, which offer expressive goal specifications while eliminating the need for expensive text-gameplay annotations. A new learning framework is derived to allow learning such instruction-following controllers from gameplay videos while producing a video instruction encoder that induces a structured goal space. We implement our agent GROOT in a simple yet effective encoder-decoder architecture based on causal transformers. We evaluate GROOT against open-world counterparts and human players on a proposed Minecraft SkillForge benchmark. The Elo ratings clearly show that GROOT is closing the human-machine gap as well as exhibiting a 70% winning rate over the best generalist agent baseline. Qualitative analysis of the induced goal space further demonstrates some interesting emergent properties, including the goal composition and complex gameplay behavior synthesis.

GROOT: Learning to Follow Instructions by Watching Gameplay Videos

2310.08235

https://openreview.net/forum?id=uleDLeiaT3

Poster

Graph Neural Networks (GNNs) are neural models that leverage the dependency structure in graphical data via message passing among the graph nodes. GNNs have emerged as pivotal architectures in analyzing graph-structured data, and their expansive application in sensitive domains requires a comprehensive understanding of their decision-making processes --- necessitating a framework for GNN explainability. An explanation function for GNNs takes a pre-trained GNN along with a graph as input, to produce a `sufficient statistic' subgraph with respect to the graph label. A main challenge in studying GNN explainability is to provide fidelity measures that evaluate the performance of these explanation functions. This paper studies this foundational challenge, spotlighting the inherent limitations of prevailing fidelity metrics, including $Fid_+$, $Fid_-$, and $Fid_\Delta$. Specifically, a formal, information-theoretic definition of explainability is introduced and it is shown that existing metrics often fail to align with this definition across various statistical scenarios. The reason is due to potential distribution shifts when subgraphs are removed in computing these fidelity measures. Subsequently, a robust class of fidelity measures are introduced, and it is shown analytically that they are resilient to distribution shift issues and are applicable in a wide range of scenarios. Extensive empirical analysis on both synthetic and real datasets are provided to illustrate that the proposed metrics are more coherent with gold standard metrics.

Towards Robust Fidelity for Evaluating Explainability of Graph Neural Networks

2310.01820

https://openreview.net/forum?id=up6hr4hIQH

Poster

Inspired by the outstanding zero-shot capability of vision language models (VLMs) in image classification tasks, open-vocabulary object detection has attracted increasing interest by distilling the broad VLM knowledge into detector training. However, most existing open-vocabulary detectors learn by aligning region embeddings with categorical labels (e.g., bicycle) only, disregarding the capability of VLMs on aligning visual embeddings with fine-grained text descriptions of object parts (e.g., pedals and bells). This paper presents DVDet, a Descriptor-Enhanced Open Vocabulary Detector that introduces conditional context prompts and hierarchical textual descriptors that enable precise region-text alignment as well as open-vocabulary detection training in general. Specifically, the conditional context prompt transforms regional embeddings into image-like representations that can be directly integrated into general open vocabulary detection training. In addition, we introduce large language models as an interactive and implicit knowledge repository which enables iterative mining and refining visually oriented textual descriptors for precise region-text alignment. Extensive experiments over multiple large-scale benchmarks show that DVDet outperforms the state-of-the-art consistently by large margins.

LLMs Meet VLMs: Boost Open Vocabulary Object Detection with Fine-grained Descriptors

2402.04630

https://openreview.net/forum?id=usrChqw6yK