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

ModeVent

Benchmarks

Task NameDataset NameSOTA ResultTrend
Hallucination EvaluationModeVent Out-of-Domain irrelevance Outliers 1.0
Average Fidelity Score4.35
42
Video Hallucination EvaluationModeVent (Overall)
Fidelity4.6
40
Video Hallucination EvaluationModeVent (Inliers)
Fidelity (F)4.8
40
Hallucination EvaluationModeVent Inliers In-Domain interference 1.0
Attribute Hijacking Fidelity2.4
4
Showing 4 of 4 rows