SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2504.18428 · NeurIPS · 2025

PolyMath: Evaluating Mathematical Reasoning in Multilingual Contexts

Rui Wang, Fei Huang, Junyang Lin, Jingren Zhou, Haoran Wei, Baosong Yang, Yiming Wang, Pei Zhang, Yichang Zhang, Jialong Tang, Chenshu Sun, Feitong Sun, and 3 more

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 1 functions out of this paper's own repositories and ran 1 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
QwenLM/PolyMath canonical 1 of 1
FunctionStatusWhere it lives
extract_boxed_content Ran QwenLM/PolyMath/eval/run_eval.py
pointer only (licence: NONE) · get_code("0db9401e2d8d130e")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

In this paper, we introduce PolyMath, a multilingual mathematical reasoning benchmark covering 18 languages and 4 easy-to-hard difficulty levels. Our benchmark ensures difficulty comprehensiveness, language diversity, and high-quality translation, making it a highly discriminative multilingual mathematical benchmark in the era of reasoning LLMs. We conduct a comprehensive evaluation for advanced LLMs and find that even Qwen3-235B-A22B-Thinking and Gemini-2.5-pro, achieve only 54.6 and 52.2 benchmark scores, with about 40% accuracy under the highest level. From a language perspective, our benchmark reveals several key challenges of LLMs in multilingual reasoning: (1) Reasoning performance varies widely across languages for current LLMs; (2) Input-output language consistency is low in reasoning LLMs and may be correlated with performance; (3) The thinking length differs significantly by language for current LLMs. Additionally, we demonstrate that controlling the output language in the instructions has the potential to affect reasoning performance, especially for some low-resource languages, suggesting a promising direction for improving multilingual capabilities in LLMs.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2504.18428")
get_code_for_paper("2504.18428")
have("2504.18428")

Connect an agent — have() is free.