CRITICTOOL: Evaluating Self-Critique Capabilities of Large Language Models in Tool-Calling Error Scenarios

2025-06-11

AI Summary

Large language models (LLMs) frequently encounter diverse errors when invoking external tools for complex tasks, yet their capabilities in error identification, diagnosis, and recovery remain poorly understood and lack systematic evaluation. Method: We introduce the first benchmark dedicated to evaluating LLMs’ self-critical capabilities for tool-call errors. It features an evolutionary data synthesis strategy to generate realistic, diverse, and multi-complexity error instances; integrates multi-dimensional error injection with fine-grained annotation of critical reasoning behaviors; and enables standardized, scalable assessment across the three stages—identification, diagnosis, and recovery. Contribution/Results: Experiments reveal that state-of-the-art models achieve less than 40% diagnostic accuracy, exposing critical weaknesses. The benchmark and evaluation framework are publicly released, establishing a new paradigm for studying reflective mechanisms in tool-augmented LLMs.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

The ability of large language models (LLMs) to utilize external tools has enabled them to tackle an increasingly diverse range of tasks. However, as the tasks become more complex and long-horizon, the intricate tool utilization process may trigger various unexpected errors. Therefore, how to effectively handle such errors, including identifying, diagnosing, and recovering from them, has emerged as a key research direction for advancing tool learning. In this work, we first extensively analyze the types of errors encountered during the function-calling process on several competitive tool evaluation benchmarks. Based on it, we introduce CRITICTOOL, a comprehensive critique evaluation benchmark specialized for tool learning. Building upon a novel evolutionary strategy for dataset construction, CRITICTOOL holds diverse tool-use errors with varying complexities, which better reflects real-world scenarios. We conduct extensive experiments on CRITICTOOL, and validate the generalization and effectiveness of our constructed benchmark strategy. We also provide an in-depth analysis of the tool reflection ability on various LLMs, offering a new perspective on the field of tool learning in LLMs. The code is available at href{https://github.com/Shellorley0513/CriticTool}{https://github.com/Shellorley0513/CriticTool}.

Problem

• Evaluating LLMs' self-critique in tool-calling error scenarios
• Analyzing error types in function-calling processes
• Assessing tool reflection ability across various LLMs

Innovation

• Evolutionary strategy for diverse dataset construction
• Comprehensive critique evaluation benchmark CRITICTOOL
• In-depth analysis of tool reflection ability