Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

KnowAgent: Knowledge-Augmented Planning for LLM-Based Agents

About

Large Language Models (LLMs) have demonstrated great potential in complex reasoning tasks, yet they fall short when tackling more sophisticated challenges, especially when interacting with environments through generating executable actions. This inadequacy primarily stems from the lack of built-in action knowledge in language agents, which fails to effectively guide the planning trajectories during task solving and results in planning hallucination. To address this issue, we introduce KnowAgent, a novel approach designed to enhance the planning capabilities of LLMs by incorporating explicit action knowledge. Specifically, KnowAgent employs an action knowledge base and a knowledgeable self-learning strategy to constrain the action path during planning, enabling more reasonable trajectory synthesis, and thereby enhancing the planning performance of language agents. Experimental results on HotpotQA and ALFWorld based on various backbone models demonstrate that KnowAgent can achieve comparable or superior performance to existing baselines. Further analysis indicates the effectiveness of KnowAgent in terms of planning hallucinations mitigation. Code is available in https://github.com/zjunlp/KnowAgent.

Yuqi Zhu, Shuofei Qiao, Yixin Ou, Shumin Deng, Shiwei Lyu, Yue Shen, Lei Liang, Jinjie Gu, Huajun Chen, Ningyu Zhang• 2024

Related benchmarks

TaskDatasetResultRank
Interactive Environment Task CompletionALFWorld Seen
Average Reward80
31
Interactive Environment Task CompletionALFWorld Unseen
Average Reward74.9
31
Embodied agentAlfWorld
Success Rate75.37
31
Interactive Environment Task CompletionScienceWorld Seen
Average Reward81.7
22
Interactive Environment Task CompletionScienceWorld Unseen
Average Reward69.6
22
Interactive Environment Task CompletionWebShop (Seen)
Average Reward64.8
22
Showing 6 of 6 rows

Other info

Follow for update