arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2409.00557cs.CLcs.AIcs.SE
收藏

学习提问:当LLM代理遇见模糊指令

Learning to Ask: When LLM Agents Meet Unclear Instruction

Renmin University of China(中国人民大学);The Chinese University of Hong Kong(香港中文大学);The Chinese University of Hong Kong, Shenzhen(香港中文大学深圳校区);Johns Hopkins University(约翰霍普金斯大学);Xiaohongshu Inc.(小红书公司)

Wenxuan Wang, Juluan Shi, Zixuan Ling, Yuk-Kit Chan, Chaozheng Wang, Cheryl Lee, Youliang Yuan, Jen-tse Huang, Wenxiang Jiao, Michael R. Lyu

AI总结:

本文提出AwN框架,通过让LLM在遇到模糊指令时主动提问以提升工具使用性能,并设计自动化评估工具ToolEvaluator,验证了在NoisyToolBench基准下的优越性。

摘要:

具备调用函数能力的现代大语言模型(LLMs)可通过外部工具解决语言技能无法完成的任务。然而,这些工具的有效执行不仅依赖于LLM的先进能力,还依赖于精确的用户指令,而现实中往往无法保证。为评估LLM在不完美指令下的工具使用性能,我们细致分析了用户查询的真实指令,识别错误模式,并构建了具有挑战性的工具使用基准NoisyToolBench。我们发现由于next-token预测训练目标,LLM倾向于任意生成缺失的参数,这可能导致幻觉和风险。为此,我们提出新颖的Ask-when-Needed(AwN)框架,促使LLM在遇到因模糊指令导致的障碍时主动向用户提问。此外,为减少用户与LLM交互的人工劳动并从准确性和效率角度评估LLM的工具使用性能,我们设计了自动化评估工具ToolEvaluator。实验表明,AwN在NoisyToolBench上显著优于现有工具学习框架。我们将释放所有相关代码和数据集以支持未来研究。

英文摘要:

Equipped with the capability to call functions, modern large language models (LLMs) can leverage external tools for addressing a range of tasks unattainable through language skills alone. However, the effective execution of these tools relies heavily not just on the advanced capabilities of LLMs but also on precise user instructions, which often cannot be ensured in the real world. To evaluate the performance of LLMs tool-use under imperfect instructions, we meticulously examine the real-world instructions queried from users, analyze the error patterns, and build a challenging tool-use benchmark called Noisy ToolBench (NoisyToolBench). We find that due to the next-token prediction training objective, LLMs tend to arbitrarily generate the missed argument, which may lead to hallucinations and risks. To address this issue, we propose a novel framework, Ask-when-Needed (AwN), which prompts LLMs to ask questions to users whenever they encounter obstacles due to unclear instructions. Moreover, to reduce the manual labor involved in user-LLM interaction and assess LLMs performance in tool utilization from both accuracy and efficiency perspectives, we design an automated evaluation tool named ToolEvaluator. Our experiments demonstrate that the AwN significantly outperforms existing frameworks for tool learning in the NoisyToolBench. We will release all related code and datasets to support future research.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑