Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

HalBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Hallucination EvaluationHalBench
Score76.7
25
Generative Hallucination MitigationObject HalBench (test)
CHAIR_S49.6
18
Showing 2 of 2 rows