Learning to Ask: When LLM Agents Meet Unclear Instruction
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911629499695104 |
|---|---|
| author | Wang, Wenxuan Shi, Juluan Ling, Zixuan Chan, Yuk-Kit Wang, Chaozheng Lee, Cheryl Yuan, Youliang Huang, Jen-tse Jiao, Wenxiang Lyu, Michael R. |
| author_facet | Wang, Wenxuan Shi, Juluan Ling, Zixuan Chan, Yuk-Kit Wang, Chaozheng Lee, Cheryl Yuan, Youliang Huang, Jen-tse Jiao, Wenxiang Lyu, Michael R. |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_00557 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Learning to Ask: When LLM Agents Meet Unclear Instruction Wang, Wenxuan Shi, Juluan Ling, Zixuan Chan, Yuk-Kit Wang, Chaozheng Lee, Cheryl Yuan, Youliang Huang, Jen-tse Jiao, Wenxiang Lyu, Michael R. Computation and Language Artificial Intelligence Software Engineering 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. |
| title | Learning to Ask: When LLM Agents Meet Unclear Instruction |
| topic | Computation and Language Artificial Intelligence Software Engineering |
| url | https://arxiv.org/abs/2409.00557 |