RealMath: A Continuous Benchmark for Evaluating Language Models on Research-Level Mathematics

Jie Zhang, Cezara Petrui, Kristina Nikolić, Florian Tramèr

Keywords: Mathematics, LLM reasoning

TL;DR: We introduce RealMath, a novel benchmark derived directly from research papers and mathematical forums that assesses LLMs' abilities on authentic mathematical tasks.

Abstract: Existing benchmarks for evaluating mathematical reasoning in large language models (LLMs) rely primarily on competition problems, formal proofs, or artificially challenging questions---failing to capture the nature of mathematics encountered in actual research environments. We introduce \textsc{RealMath}, a novel benchmark derived directly from research papers and mathematical forums that assesses LLMs' abilities on authentic mathematical tasks. Our approach addresses three critical challenges: sourcing diverse research-level content, enabling reliable automated evaluation through verifiable statements, and designing a continually refreshable dataset to mitigate contamination risks. Experimental results across multiple LLMs reveal surprising capabilities in handling research mathematics compared to competition problems, suggesting current models may already serve as valuable assistants for working mathematicians despite limitations on highly challenging problems. The code and dataset for \textsc{RealMath} are publicly available.

Dataset URL: https://huggingface.co/datasets/ethz-spylab/RealMath

Code URL: https://github.com/ethz-spylab/RealMath

Primary Area: Evaluation (e.g., data collection methodology, data processing methodology, data analysis methodology, meta studies on data sources, extracting signals from data, replicability of data collection and data analysis and validity of metrics, validity of data collection experiments, human-in-the-loop for data collection, human-in-the-loop for data evaluation)

Submission Number: 1288