---
title: README
emoji: 🦀
colorFrom: blue
colorTo: blue
sdk: static
pinned: false
---
## Model/Data associated with research project *Autonomous Evaluation and Refinement of Digital Agents*.

### [Paper](https://arxiv.org/abs/2404.06474) | [Code](https://github.com/Berkeley-NLP/Agent-Eval-Refine)


We design and use model-based evaluators to both evaluate and autonomously refine the performance of digital agents. Experiments show that domain-general automated evaluators can significantly improve the performance of digital agents, without any extra supervision.


[Jiayi Pan](https://www.jiayipan.me/), [Yichi Zhang](https://sled.eecs.umich.edu/author/yichi-zhang/), [Nicholas Tomlin](https://people.eecs.berkeley.edu/~nicholas_tomlin/), [Yifei Zhou](https://yifeizhou02.github.io/), [Sergey Levine](https://people.eecs.berkeley.edu/~svlevine/), [Alane Suhr](https://www.alanesuhr.com/)

UC Berkeley, University of Michigan