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

Hallucination Detection research benchmarks – Page 13 · SOTA2 Research

Hallucination Detection

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

87AUROC

7

May 7, 2026

GPT-4o (teacher)

94.1Precision

7

Apr 28, 2026

Earth Observation

90.94F1 Score

7

Apr 16, 2026

MHaluBench Image-to-Text Segment-level

Gemini-based Self-Check

90.44Hallucinatory Precision

7

Apr 7, 2026

MHaluBench Image-to-Text (Claim-level)

GPT-based Self-Check

86.54Hallucinatory Precision

7

Apr 7, 2026

MetaQA 1hop (LLaMA2-7B)

GGA (PRD + SAS)

83.41AUC

7

Feb 26, 2026

HaluEval Summarization (Starling-LM-7B-alpha)

81TPR

7

Feb 26, 2026

76.8Algorithmic

7

Feb 26, 2026

OpenDialKG Eval (test)

76.2Macro F1

7

Feb 26, 2026

1,800Detection Percentage

7

Feb 26, 2026

HalOmi Zero-Shot 1.0

0.66ROC AUC

7

Feb 26, 2026

HalOmi Low-Resource 1.0

0.79ROC AUC

7

Feb 26, 2026

HalOmi High-Resource 1.0

0.91AUC (ROC)

7

Feb 26, 2026

Maynez Factual 2020

0.264Spearman Correlation

7

Feb 26, 2026

0.431Spearman Correlation

7

Feb 26, 2026

WMT DE-EN 18 (test)

91.72AUROC

7

Feb 26, 2026

MLQE-PE NE-EN (test)

97.64AUROC

7

Feb 26, 2026

MLQE-PE RO-EN (test)

99.72AUROC

7

Feb 26, 2026

wiki_bio gpt3_hallucination

0.751ROC AUC

6

Jul 9, 2026

KG-Guard(GraphTransformer)

0LLM Calls

6

Jun 2, 2026

KG-Guard(GraphTransformer)

0LLM Calls

6

Jun 2, 2026

KG-Guard(GraphTransformer)

0LLM Calls

6

Jun 2, 2026

88.89EigenWD M1

6

Mar 25, 2026

64.8EigenWD M1 Score

6

Mar 25, 2026

Overall CovidQA, PubmedQA, FinanceBench

82.3F1 Score

6

Mar 23, 2026

Page 13 of 17

Popular datasets

RAGTruth HaluEval TriviaQA SQuAD CoQA

Follow for updates

Join SOTA2 Contact