{c}
{tag=AI Math benchmark}
{title2=2025}
{wiki}

Contains highly specialized questions in various academic fields, including <mathematics>. The problems are answered either with a number, or multiple choice, or free text.

* https://arxiv.org/abs/2501.1424
* https://huggingface.co/datasets/cais/hle
* https://agi.safe.ai/


Humanity's Last Exam

{c}
{tag=Computer benchmark}

This section is about benchmarks designed to test mathematical reasoning.

Bibliography:
* https://mathscholar.org/2025/02/deepseek-a-breakthrough-in-ai-for-math-and-everything-else/


Ciro Santilli @cirosantilli 37

 Tagged: AI Math benchmark