Object Hallucination Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
88.57Random Accuracy
8
Feb 26, 2026
89.24Accuracy (Random)
6
Jun 30, 2026
92.6Rand Accuracy
6
Jun 30, 2026
76.6Accuracy
6
Jun 16, 2026
437Latency (10/token) (ms)
5
Jun 29, 2026
91.47F1 Score
5
May 15, 2026
675VLM Judgment
5
Feb 26, 2026
82.8Score (Noise)
4
Apr 24, 2026
26Total Time
4
Apr 14, 2026
77.67Score (Zh)
4
Feb 26, 2026
86.5Zh Score
4
Feb 26, 2026
77.87Adversarial Score
3
Jun 18, 2026
86.4Accuracy
3
Jun 16, 2026
72.4Top-1 Accuracy
3
Feb 26, 2026
75.2Top-1 Accuracy
3
Feb 26, 2026
88.3Top-1 Accuracy
3
Feb 26, 2026
9.6CHAIR Score
3
Feb 26, 2026
81.07Accuracy
2
May 15, 2026
86.67Accuracy
2
May 15, 2026
91.23Accuracy
2
May 15, 2026
7.28Accuracy
2
Apr 2, 2026