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

Open Question Answering research benchmarks · SOTA2 Research

Open Question Answering

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Natural Questions (NQ) (test)

58.4Exact Match (EM)

134

Feb 26, 2026

WEBQUESTIONS (test)

Memory Networks

42.2F1 Score

27

Feb 26, 2026

Llama3.3-LightRAG-mix

98.11SS Score

10

Jun 2, 2026

AIME 2025 (test)

70Accuracy

9

Feb 26, 2026

0.94Accuracy

9

Feb 26, 2026

88.2Accuracy

8

Mar 4, 2026

76.4Factual Correctness (%)

8

Feb 26, 2026

LegalMC4 (test)

GPT-5 (min. reasoning)

77.2LLM Factual Correctness

8

Feb 26, 2026

54.31FENSE Score

3

Mar 27, 2026

62.35FENSE Score

3

Mar 27, 2026

Popular datasets

AIME AudioCaps Natural Questions WEBQUESTIONS OpenQA

Follow for updates

Join SOTA2 Contact