Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Hallucination Mitigation on POPE (val)
Loading...
88.2
ACC
VIB-Probe
81.96
83.58
85.2
86.82
Jan 9, 2026
ACC
F1
Updated 4d ago
Evaluation Results
Method
Method
Links
ACC
F1
VIB-Probe
Base Model=LLaVA-v1.6-7B
2026.01
88.2
89.5
PAI
Base Model=LLaVA-v1.6-7B
2026.01
87.9
88.4
VCD
Base Model=LLaVA-v1.6-7B
2026.01
86.3
87.8
BeamSearch
Base Model=LLaVA-v1.6-7B
2026.01
84.3
85.6
Vanilla
Base Model=LLaVA-v1.6-7B
2026.01
84.1
85.1
PAI
Base Model=LLaVA-v1.5-7B
2026.01
84
84.6
VIB-Probe
Base Model=LLaVA-v1.5-7B
2026.01
83.7
85.2
VCD
Base Model=LLaVA-v1.5-7B
2026.01
83.6
83.9
Vanilla
Base Model=LLaVA-v1.5-7B
2026.01
82.6
83.3
BeamSearch
Base Model=LLaVA-v1.5-7B
2026.01
82.2
84.1
Feedback
Search any
task
Search any
task