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

MuSiQue dataset and benchmarks – Page 10 · SOTA2 Research

Research Datasets

MuSiQue

Benchmarks

Task NameDataset NameSOTA MethodSort

Task NameDataset NameSOTA ResultTrendResults

Multi-hop Question Answering

MuSiQue

23.6LJFT

5

Short-form Question Answering

Musique (test)

13.67EM

5

Multi-step Question Answering

MuSiQue (test)

53.8Exact Match

5

Open-domain Question Answering

MuSiQue (test)

26.5EM

5

MuSiQue

42.57Recall@2

4

Knowledge-intensive Question Answering

MuSiQue

2.9Truthfulness

4

Error Recognition

Musique

38.9Accuracy@1

4

MuSiQue

73Cumulative Gold-Document Recall

4

MuSiQue

86.4Recall@10

4

Tool-augmented Question Answering

Musique

17.33Accuracy

4

Open-Domain Question Answering

MuSiQue

3.8Joint Answer-and-Evidence Score

4

Multi-hop Question Answering

MuSiQue

10.3Evidence Score

4

Question Answering

Musique

18.6Mean Score (Musique QA)

4

Multi-step reasoning

Musique 16k ~ 20k

29.5Macro Accuracy

4

Multi-step reasoning

Musique 12k ~ 16k

31.6Accuracy (macro)

4

Multi-step reasoning

Musique 8k ~ 12k

34.3Macro Accuracy

4

MuSiQue

0.83Avg Violation

4

Question Answering

MuSiQue

1.988R@1

4

Multi-hop Question Answering

MuSiQue Full (dev)

33.64EM Score

4

Multi-hop Question Answering

MuSiQue

38.5EM

4

Open-domain Question Answering

MuSiQue 500 randomly chosen questions

17.6EM

4

Open-domain Question Answering

MuSiQue (500 randomly chosen questions)

1BM25 Queries

4

Question Answering

MuSiQue

1.9Indexing Time (min)

4

Tool Attribution Correctness

Musique

86.5Precision

4

Text-only Question Answering

MuSiQue

11.5EM

4

Page 10 of 12

Popular tasks

Multi-hop Question Answering Question Answering Multi-hop Reasoning Open-domain Question Answering Retrieval

Follow for updates

Join SOTA2 Contact