Adaptive Tool Use in Large Language Models with Meta-Cognition Trigger
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912547028860928 |
|---|---|
| author | Li, Wenjun Li, Dexun Dong, Kuicai Zhang, Cong Zhang, Hao Liu, Weiwen Wang, Yasheng Tang, Ruiming Liu, Yong |
| author_facet | Li, Wenjun Li, Dexun Dong, Kuicai Zhang, Cong Zhang, Hao Liu, Weiwen Wang, Yasheng Tang, Ruiming Liu, Yong |
| contents | Large language models (LLMs) have shown remarkable emergent capabilities, transforming the execution of functional tasks by leveraging external tools for complex problems that require specialized processing or up-to-date data. While existing research expands LLMs access to diverse tools (e.g., program interpreters, search engines, calculators), the necessity of using these tools is often overlooked, leading to indiscriminate tool invocation. This naive approach raises two key issues: increased latency due to unnecessary tool calls, and potential errors resulting from faulty interactions with external tools. In this paper, we introduce meta-cognition as a proxy for LLMs self-assessment of their capabilities, reflecting the model's awareness of its own limitations. Based on this, we propose MeCo, an adaptive decision-making strategy for external tool use. MeCo quantifies metacognitive scores by capturing high-level cognitive signals in the representation space, guiding when to invoke tools. Notably, MeCo is fine-tuning-free and incurs minimal cost. Experiments across multiple backbone models and benchmarks show that MeCo reliably detects LLMs' internal cognitive signals and significantly improves tool-use decision-making. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_12961 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Adaptive Tool Use in Large Language Models with Meta-Cognition Trigger Li, Wenjun Li, Dexun Dong, Kuicai Zhang, Cong Zhang, Hao Liu, Weiwen Wang, Yasheng Tang, Ruiming Liu, Yong Computation and Language Artificial Intelligence Large language models (LLMs) have shown remarkable emergent capabilities, transforming the execution of functional tasks by leveraging external tools for complex problems that require specialized processing or up-to-date data. While existing research expands LLMs access to diverse tools (e.g., program interpreters, search engines, calculators), the necessity of using these tools is often overlooked, leading to indiscriminate tool invocation. This naive approach raises two key issues: increased latency due to unnecessary tool calls, and potential errors resulting from faulty interactions with external tools. In this paper, we introduce meta-cognition as a proxy for LLMs self-assessment of their capabilities, reflecting the model's awareness of its own limitations. Based on this, we propose MeCo, an adaptive decision-making strategy for external tool use. MeCo quantifies metacognitive scores by capturing high-level cognitive signals in the representation space, guiding when to invoke tools. Notably, MeCo is fine-tuning-free and incurs minimal cost. Experiments across multiple backbone models and benchmarks show that MeCo reliably detects LLMs' internal cognitive signals and significantly improves tool-use decision-making. |
| title | Adaptive Tool Use in Large Language Models with Meta-Cognition Trigger |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2502.12961 |