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

General Evaluation research benchmarks · SOTA2 Research

General Evaluation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Aggregate Benchmarks

Gemini 2.5 Pro+ASP

93.9Average Score

37

May 13, 2026

DeepSeek-R1 (BF16)

70.22Accuracy

29

Mar 17, 2026

Average Across all benchmarks

6.9Speedup

28

Jun 2, 2026

UltraFeedback Aggregate

59.59Overall Average Score

18

Jul 7, 2026

46.83Accuracy

15

May 18, 2026

53.74Overall Average Score

12

Jun 2, 2026

39.5REC

12

Feb 26, 2026

Aggregate Suite PIQA, HellaSwag, WinoGrande, ARC-e, ARC-c

69Average Score

10

Feb 26, 2026

Reasoning, Knowledge, and Biomedicine combined datasets (test)

60.47Average Score

9

Feb 27, 2026

71.8Score

8

Jun 2, 2026

Downstream Suite

39.38Average Score

8

May 18, 2026

DeepLatent-RL-7B

86.4Score

7

Jun 2, 2026

DeepLatent-RL-7B

38.7Score

7

Jun 2, 2026

Average Downstream Benchmark Suite

37.9Average Accuracy

7

Apr 10, 2026

Datacomp small (38 tasks)

Final Self-Filtered 30% Data Mix

19.7Average Score

6

Jun 23, 2026

General Teacher

52.1Score

6

May 27, 2026

Instruction Tuning Suite (BIG-bench Hard, MMLU, TyDi QA, MGSM)

Flan-PaLM 2 (L)

74.1Average Score

4

Feb 26, 2026

ExpSuite-Static Overall

78.75Average Score

2

Jun 1, 2026

Popular datasets

LiveBench MM-Vet AGIEval UltraFeedback ChartQA

Follow for updates

Join SOTA2 Contact