UI-Evol: Automatic Knowledge Evolving for Computer Use Agents

TL;DR

UI-Evol is a modular system that enhances computer use agents by evolving GUI knowledge through interaction data and external references, significantly improving task success rates and reliability.

Contribution

It introduces a novel plug-and-play knowledge evolution module with retrace and critique stages, addressing the knowledge-execution gap in computer use agents.

Findings

UI-Evol significantly improves task performance on OSWorld benchmark.

It reduces behavioral standard deviation, increasing agent reliability.

Demonstrates effectiveness over state-of-the-art Agent S2.

Abstract

External knowledge has played a crucial role in the recent development of computer use agents. We identify a critical knowledge-execution gap: retrieved knowledge often fails to translate into effective real-world task execution. Our analysis shows even 90% correct knowledge yields only 41% execution success rate. To bridge this gap, we propose UI-Evol, a plug-and-play module for autonomous GUI knowledge evolution. UI-Evol consists of two stages: a Retrace Stage that extracts faithful objective action sequences from actual agent-environment interactions, and a Critique Stage that refines existing knowledge by comparing these sequences against external references. We conduct comprehensive experiments on the OSWorld benchmark with the state-of-the-art Agent S2. Our results demonstrate that UI-Evol not only significantly boosts task performance but also addresses a previously overlooked issue of high behavioral standard deviation in computer use agents, leading to superior performance on computer use tasks and substantially improved agent reliability.