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 139 · SOTA2 Research

Mathematical Reasoning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

StructU+Self-ConsU

92.5Sel-AUC

3

Jun 17, 2026

26Pass@100

3

Jun 12, 2026

53.3Pass@1

3

Jun 8, 2026

AIME Average context 2022-2024

0.7 × RoPE theta × 16 + 0.3 × 1M-Model

28Accuracy

3

Jun 4, 2026

AIME Long context 2022-2024

0.7 × RoPE theta × 16 + 0.3 × 1M-Model

29.33Accuracy

3

Jun 4, 2026

AIME Short context 2022-2024

0.7 × RoPE theta × 16 + 0.3 × 1M-Model

26.67Accuracy

3

Jun 4, 2026

MATH500 Long context

0.7 × RoPE theta × 16 + 0.3 × 1M-Model

89.12Accuracy

3

Jun 4, 2026

Speculative Thinking

19,376.5Average Output Length

3

Jun 4, 2026

AIME 2020–2024

Speculative Thinking

15,536.7Average Length

3

Jun 4, 2026

PyraMathBench multimodal (test)

73VRP

3

Jun 4, 2026

95.15Pass@8

3

Jun 4, 2026

16.67Pass@8

3

Jun 4, 2026

Mathematical Reasoning Average

38.66Mean@8 Accuracy

3

Jun 4, 2026

81.79Mean@8 Accuracy

3

Jun 4, 2026

39.69Mean@8 Accuracy

3

Jun 4, 2026

66.92Mean@8 Accuracy

3

Jun 4, 2026

53.9GSM8K Accuracy

3

Jun 4, 2026

5.1MT-B Score

3

Jun 4, 2026

Mathematical Reasonings Average

Qwen3-14B + GRPO

60.9Accuracy

3

Jun 1, 2026

Hard Math Dataset

86MATH Score

3

May 29, 2026

16Accuracy

3

May 29, 2026

31OlymE Accuracy

3

May 29, 2026

53.93Accuracy

3

May 29, 2026

65.4Throughput (Tokens/sec)

3

May 28, 2026

32.6Mean Score (@16)

3

May 28, 2026

Page 139 of 152

Popular datasets

AIME MATH GSM8K AMC OlympiadBench

Follow for updates

Join SOTA2 Contact