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

Overall research benchmarks · SOTA2 Research

Overall

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

78.6Overall Score

27

Jun 4, 2026

0.93Pearson Correlation (ρ)

13

Apr 23, 2026

Loong Set 4: 200K–250K Tokens

54.62LLM Score

12

Feb 26, 2026

Loong Set 3: 100K–200K Tokens

58.86LLM Score

12

Feb 26, 2026

Loong Set 2: 50K–100K Tokens

0.636LLM Score

12

Feb 26, 2026

Loong Set 1: 10K–50K Tokens

71LLM Score

12

Feb 26, 2026

Principle-based evaluation dataset

Claude 4.5-Haiku

8.41Average

12

Feb 26, 2026

5,100Success Rate

10

Feb 26, 2026

9,700Success Rate

10

Feb 26, 2026

humanoidmaze giant

79Success Rate

10

Feb 26, 2026

StethoBench (test)

71.8BERTScore

8

Mar 4, 2026

46.5ROUGE-1

8

Mar 4, 2026

PDDLLM v1 (test)

95.7Planning Success Rate

6

Feb 26, 2026

Standard Evaluation Suite

0.765Average Score

6

Feb 26, 2026

AffordPose refined (test)

95Success Rate

4

Feb 26, 2026

The benchmark held-out slice

Solver-Grounded Synthetic CoT (Model)

85Accuracy

2

Jun 23, 2026

SemEval 2017 Task 10 ScienceIE (test)

48.1Precision

2

Feb 26, 2026

Popular datasets

Loong puzzle StethoBench SemEval Principle-based evaluation

Follow for updates

Join SOTA2 Contact