Scheherazade: Evaluating Chain-of-Thought Math Reasoning in LLMs with Chain-of-Problems
This addresses the need for scalable, automated benchmark creation to better evaluate LLM reasoning capabilities, though it is incremental as it builds on existing datasets like GSM8K.
The authors tackled the problem of existing math reasoning benchmarks becoming too easy for state-of-the-art LLMs by introducing Scheherazade, an automated method to generate challenging benchmarks through logical chaining of problems, resulting in performance declines for most frontier models except OpenAI's o1-preview, which maintained better performance, especially in backward reasoning.
Benchmarks are critical for measuring Large Language Model (LLM) reasoning capabilities. Some benchmarks have even become the de facto indicator of such capabilities. However, as LLM reasoning capabilities improve, existing widely-used benchmarks such as GSM8K marginally encapsulate model reasoning differentials - most state-of-the-art models for example achieve over 94% accuracy on the GSM8K dataset (paperwithcode, 2024). While constructing harder benchmarks is possible, their creation is often manual, expensive, and unscalable. As such, we present Scheherazade, an automated approach to produce large quantities of challenging mathematical reasoning benchmarks by logically chaining a small starting set of problems. We propose two different chaining methods, forward chaining and backward chaining, which include randomized branching techniques to generate complex reasoning problems. We apply Scheherazade on GSM8K to create GSM8K-Scheherazade and evaluate 3 frontier LLMs and OpenAI's o1-preview on it. We show that while other frontier models' performance declines precipitously at only a few questions chained, our evaluation suggests o1-preview's performance persists, with the flagship OpenAI model the only one to perform better at backward reasoning. Our data and code are available at https://github.com/YoshikiTakashima/scheherazade-code-data.