Adaptive Tool Use in Large Language Models with Meta-Cognition Trigger

TL;DR

This paper introduces MeCo, a meta-cognition-based method enabling large language models to adaptively decide when to use external tools, thereby reducing unnecessary calls and errors without additional training.

Contribution

It presents a fine-tuning-free, minimal-cost meta-cognitive approach for LLMs to self-assess and improve external tool invocation decisions.

Findings

MeCo effectively detects LLMs' internal cognitive signals.

It significantly improves tool-use decision accuracy.

The method reduces unnecessary tool calls and errors.

Abstract

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.