Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

STEM Reasoning research benchmarks – Page 2 · SOTA2 Research

STEM Reasoning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

17.3Accuracy

7

May 21, 2026

60.9Score

7

Feb 26, 2026

67.6Score

7

Feb 26, 2026

Qwen3-8B-as-GenRM

77.7Score

7

Feb 26, 2026

76.9STEM Reasoning Score

6

Apr 29, 2026

MMMU-Pro Vision

Qwen-8B-DeltaThinker

56.3Accuracy

4

May 18, 2026

61.62Accuracy

3

May 20, 2026

Minerva 272 undergraduate-level STEM problems

56.57Avg@32 Score

3

Mar 24, 2026

Page 2 of 2

Popular datasets

GPQA MMLU MMMU Minerva AIME

Follow for updates

Join SOTA2 Contact