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

Question Answering research benchmarks – Page 104 · SOTA2 Research

Question Answering

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Classifier+LLM (Structured)

53.1EM

4

Feb 26, 2026

TriviaQA (test)

35.6EM

4

Feb 26, 2026

95.6HS

4

Feb 26, 2026

94.9HS

4

Feb 26, 2026

95.5HS

4

Feb 26, 2026

6Normalized PLL Score

4

Feb 26, 2026

BioStarsGPT (test)

Best Unfine-Tuned (BUT) Baseline

0Exact Match

4

Feb 26, 2026

SimpleQuestions (train/test)

35.31BLEU

4

Feb 26, 2026

FreshQA (train test)

28.37BLEU

4

Feb 26, 2026

TruthfulQA (train test)

0.523BLEU

4

Feb 26, 2026

HotpotQA (train test)

53.19BLEU

4

Feb 26, 2026

74.73Ps

4

Feb 26, 2026

AmbigQA (sampled)

65.5Accuracy

4

Feb 26, 2026

69.5Accuracy

4

Feb 26, 2026

Bamboogle (sampled)

68.8Accuracy

4

Feb 26, 2026

MuSiQue (sampled)

31.5Accuracy

4

Feb 26, 2026

2WikiMQA (sampled)

0.63Accuracy

4

Feb 26, 2026

HotpotQA (sampled)

54Accuracy

4

Feb 26, 2026

xbench DeepSearch

SearchGym-RL (Ours)

56Accuracy (Pass@4)

4

Feb 26, 2026

ChatGPT 5 Thinking

70Accuracy

4

Feb 26, 2026

99Accuracy

4

Feb 26, 2026

MS MARCO MQA (test val)

69.4F1 Score

4

Feb 26, 2026

84.3F1 Score

4

Feb 26, 2026

2WikiMultihopQA 56K context

73.4F1 Score

4

Feb 26, 2026

2WikiMultihopQA 28K context

73.4F1

4

Feb 26, 2026

Page 104 of 122

Popular datasets

HotpotQA TriviaQA ARC NQ SQuAD

Follow for updates

Join SOTA2 Contact