Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Object-HalBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Hallucination assessmentObject-HalBench
Mention Hallucination Rate2.6
39
Generative HallucinationObject HalBench
CHAIR_S Score61
33
Hallucination EvaluationObject HalBench
CHAIR Score (s)52.7
28
Hallucination EvaluationObject HalBench full benchmark
Ha (Living Room)25.2
5
Showing 4 of 4 rows