From Proof to Program: Characterizing Tool-Induced Reasoning Hallucinations in Large Language Models

ACL ARR 2026 January Submission6637 Authors

Keywords: Tool-augmented Language Models, Tool-induced Reasoning Hallucinations, Evaluation Benchmarks, Hallucination Mitigation

Abstract:

Tool-augmented Language Models (TaLMs) can invoke external tools to solve problems beyond their parametric capacity. However, it remains unclear whether these tool-enabled gains reflect trustworthy reasoning. Focusing on the Code Interpreter tool, we show that even when tools are selected and executed correctly, TaLMs treat tool outputs as substitutes for reasoning, producing solutions that appear correct but lack coherent justification. We term this failure mode Tool-Induced Myopia (TIM), and study it using PyMath, a benchmark of 1,679 competition-level mathematical problems for which Python code is helpful but not sufficient.
We further develop a multi-dimensional evaluation suite to quantify reasoning degradation in TaLMs relative to their non-tool counterparts. Our findings reveal that while TaLMs achieve up to a 19.3 percentage point gain in final-answer accuracy, their reasoning behavior consistently deteriorates (e.g., non-tool language models win up to 41.5\% more often in pairwise comparisons of reasoning processes). This degradation intensifies with tool use; the more frequently a model invokes tools, the less coherent its reasoning becomes. Moreover, tool use shifts errors from arithmetic mistakes toward global reasoning failures (logic, assumption, creativity).
Finally, we propose a preference-optimization-based framework that realigns TaLMs to use tool outputs as assistive evidence, improving both final-answer accuracy and reasoning depth under tool use. Code and data will be released upon publication.

Paper Type: Long

Research Area: AI/LLM Agents

Research Area Keywords: AI/LLM Agents, Language Modeling, Resources and Evaluation, Code Models, Generation

Contribution Types: Model analysis & interpretability, NLP engineering experiment, Data resources, Data analysis

Languages Studied: English

Submission Number: 6637