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

Mathematical Reasoning research benchmarks – Page 93 · SOTA2 Research

Mathematical Reasoning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

80.44Accuracy

10

Mar 9, 2026

43.2Accuracy

10

Mar 9, 2026

77.6Accuracy

10

Apr 2, 2026

In-domain Reasoning Suite (AIME24, AIME25, AMC23, Math500, Minerva, Olympia) (test)

29.3AIME24 Score

10

Mar 4, 2026

90GSM8K Score

10

Mar 4, 2026

52.94Correction Uplift (%)

10

Mar 4, 2026

27.59Correction Uplift

10

Mar 4, 2026

48.87Correction Uplift

10

Mar 4, 2026

OPD (prefix scheduling)

10.6Mean Accuracy

10

Feb 27, 2026

4.818LIS

10

Feb 26, 2026

Out-of-Domain Mathematical Reasoning Suite ARC-c, GPQA, MMLU-Pro

84.1ARC-c Score

10

Feb 26, 2026

29.78Exact Match Accuracy

10

Feb 26, 2026

Math Evaluation Suite

Sparse upcycling

28.1Math Score

10

Feb 26, 2026

11.1Accuracy

10

Feb 26, 2026

39.8Accuracy

10

Feb 26, 2026

81.3Accuracy

10

Feb 26, 2026

89.5avg@16

10

Feb 26, 2026

41.5Avg@16

10

Feb 26, 2026

Overall Combined Benchmarks

58.4Avg@3 Score

10

Feb 26, 2026

58.7Avg@3

10

Feb 26, 2026

97.5Accuracy (Avg@4)

10

Mar 12, 2026

74.66Accuracy

10

Feb 26, 2026

Countdown 4,5,6-arg held-out difficulties (test)

SkillFactory -> GRPO

25.1Accuracy

10

Feb 26, 2026

85.6Pass@1

10

Feb 26, 2026

66.27Pass@1

10

Feb 26, 2026

Page 93 of 152

Popular datasets

AIME MATH GSM8K AMC OlympiadBench

Follow for updates

Join SOTA2 Contact