SelfGoal: Your Language Agents Already Know How to Achieve High-level Goals
This addresses a key challenge for developers and users of language agents in domains like gaming and programming, though it appears incremental as it builds on existing agent frameworks.
The paper tackles the problem of language agents struggling to achieve high-level goals with limited instructions and delayed feedback by introducing SelfGoal, an approach that adaptively breaks goals into subgoals and updates them during interaction, resulting in significant performance improvements across competitive, cooperative, and deferred feedback tasks.
Language agents powered by large language models (LLMs) are increasingly valuable as decision-making tools in domains such as gaming and programming. However, these agents often face challenges in achieving high-level goals without detailed instructions and in adapting to environments where feedback is delayed. In this paper, we present SelfGoal, a novel automatic approach designed to enhance agents' capabilities to achieve high-level goals with limited human prior and environmental feedback. The core concept of SelfGoal involves adaptively breaking down a high-level goal into a tree structure of more practical subgoals during the interaction with environments while identifying the most useful subgoals and progressively updating this structure. Experimental results demonstrate that SelfGoal significantly enhances the performance of language agents across various tasks, including competitive, cooperative, and deferred feedback environments. Project page: https://selfgoal-agent.github.io.