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

Error detection research benchmarks – Page 3 · SOTA2 Research

Error detection

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

69.8Precision

6

Feb 26, 2026

0.974Precision

6

Feb 26, 2026

100Precision

6

Feb 26, 2026

99.8Precision

6

Feb 26, 2026

MAQA* High-Ambiguity Subset H[p*] >= 1.5 (test)

65AUROC

5

Feb 26, 2026

TruthfulQA (full)

B1 mean entropy

0.599AUROC

4

Mar 26, 2026

Semantic Entropy

0.81Rho(U_epi, U_ale)

4

Feb 26, 2026

TtTSet 5662 samples (test)

55.6Accuracy

4

Feb 26, 2026

SEED IV (mean over 3 folds)

0.538AUROC

3

May 11, 2026

PhysionetMI (mean over 3 folds)

0.764AUROC

3

May 11, 2026

BCI-IV-2a (mean over 3 folds)

0.636AUROC

3

May 11, 2026

TruthfulQA 200-question subset

0.512AUROC

3

Mar 26, 2026

Combined Error Detection Spreadsheet-Tables, Relational-Tables

Table-Specialist

33.2Quality Score

3

Mar 25, 2026

97.7F1 Score (Correct)

3

Mar 5, 2026

94.4Correct (F1)

3

Mar 5, 2026

NeckData 5-fold cross-participant

65.8Accuracy

3

Feb 26, 2026

66.8Precision

3

Feb 26, 2026

67.2Noun EDA

3

Feb 26, 2026

OpenData 5-fold cross-participant

60.6Accuracy

2

Feb 26, 2026

72.53Precision

2

Feb 26, 2026

bi-directional LSTM with auxiliary objectives

45.1F0.5

2

Feb 26, 2026

bi-directional LSTM with auxiliary objectives

36.1F0.5 Score

2

Feb 26, 2026

bi-directional LSTM with auxiliary objectives

61.2F0.5 Score

2

Feb 26, 2026

RGB Frame-based 5-fold cross-participant

51.5Accuracy

1

Feb 26, 2026

NeckData Single-participant

84.7Accuracy

1

Feb 26, 2026

Page 3 of 4

Popular datasets

ImageNet Conll HotpotQA MuSiQue TruthfulQA

Follow for updates

Join SOTA2 Contact