Is Your Model Really A Good Math Reasoner? Evaluating Mathematical Reasoning with Checklist
This addresses the critical issue of overfitting in mathematical reasoning benchmarks for AI researchers and developers by providing a more comprehensive evaluation method.
The authors tackled the problem of comprehensively evaluating mathematical reasoning abilities in large language models by introducing MathCheck, a checklist-based evaluation framework that tests task generalization and reasoning robustness. Their results showed that while frontier models like GPT-4o performed well, many other models exhibited significant performance declines on MathCheck compared to traditional benchmarks.
Exceptional mathematical reasoning ability is one of the key features that demonstrate the power of large language models (LLMs). How to comprehensively define and evaluate the mathematical abilities of LLMs, and even reflect the user experience in real-world scenarios, has emerged as a critical issue. Current benchmarks predominantly concentrate on problem-solving capabilities, presenting a substantial risk of model overfitting and fails to accurately measure the genuine mathematical reasoning abilities. In this paper, we argue that if a model really understands a problem, it should be robustly applied across a diverse array of tasks. To this end, we introduce MathCheck, a well-designed checklist for testing task generalization and reasoning robustness, as well as an automatic tool to generate checklists efficiently. MathCheck includes multiple mathematical reasoning tasks and robustness tests to facilitate a comprehensive evaluation of both mathematical reasoning ability and behavior testing. Utilizing MathCheck, we develop MathCheck-GSM and MathCheck-GEO to assess math textual reasoning and multi-modal reasoning abilities, respectively, serving as upgraded versions of benchmarks including GSM8k, GeoQA, UniGeo, and Geometry3K. We adopt MathCheck-GSM and MathCheck-GEO to evaluate 26 LLMs and 17 MLLMs. Our results demonstrate that while frontier LLMs like GPT-4o continue to excel in various abilities on the checklist, many other model families exhibit a significant decline. Further experiments indicate that, compared to traditional math benchmarks, MathCheck better reflects true mathematical abilities and represents mathematical intelligence more linearly, thereby supporting our design. Using MathCheck, we can efficiently conduct informative behavior analysis to deeply investigate models. Finally, we show that our checklist paradigm can easily extend to other reasoning tasks.