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 Generation research benchmarks – Page 2 · SOTA2 Research

Question Generation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

24.59BLEU-4

6

Feb 26, 2026

HotpotQA (test)

0.211BLEU-3

6

Feb 26, 2026

Physics Word Problem Dataset

Llama-3.2-3B-Instruct

58.73Conformity Score

5

Jun 16, 2026

TextRS-300 (test)

44.26BLEU-1

5

Feb 26, 2026

NWPU-300 (test)

41.87BLEU-1

5

Feb 26, 2026

53Grammaticality Wins

5

Feb 26, 2026

TriviaQA (test)

35.45QAE EM

5

Feb 26, 2026

Natural Questions (test)

37.18QAE EM

5

Feb 26, 2026

Molweni All-possible Names

19.57BLEU

4

Feb 26, 2026

Molweni In-distribution Names

19.58BLEU

4

Feb 26, 2026

Human Evaluation 2-hop questions (test)

74Well-formed Rate (Yes)

4

Feb 26, 2026

2.91Grammar

3

Jun 5, 2026

1.95Fluency

3

Mar 24, 2026

44.26BLEU-1

3

Feb 26, 2026

41.87BLEU-1

3

Feb 26, 2026

WebQuestions (test)

4.53Syntactic Score

3

Feb 26, 2026

OR-ShARC unseen (test)

34.9F1 BLEU-1

3

Feb 26, 2026

OR-ShARC (test seen)

83.4F1-BLEU-1

3

Feb 26, 2026

Jeopardy (test)

17.3B-1

3

Feb 26, 2026

49.7Diversity Win

3

Feb 26, 2026

29.15ROUGE-1

2

Feb 26, 2026

45.26ROUGE-1

2

Feb 26, 2026

Baseline QG Model

4.83Fluency Score

2

Feb 26, 2026

Human Evaluation 1-hop questions (test)

56Well-formed: Yes

2

Feb 26, 2026

42.7Factuality

2

Feb 26, 2026

Page 2 of 3

Popular datasets

SQuAD OR-ShARC Molweni Human Evaluation Jeopardy

Follow for updates

Join SOTA2 Contact