Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Membership Inference on MIA-Bench 5% Forget Set
Loading...
69.6
Average Performance
Vanilla
46.8968
52.7909
58.685
64.5791
Feb 21, 2025
Average Performance
Updated 1mo ago
Evaluation Results
Method
Method
Links
Average Performance
Vanilla
Base Model=LLaVA, Forg...
2025.02
69.6
MANU
Base Model=LLaVA, Forg...
2025.02
64.02
NPO
Base Model=LLaVA, Forg...
2025.02
63.44
KL Minim.
Base Model=LLaVA, Forg...
2025.02
62.24
Grad. Diff
Base Model=LLaVA, Forg...
2025.02
59.02
GA
Base Model=LLaVA, Forg...
2025.02
57.29
Vanilla
Base Model=Idefics2, F...
2025.02
54.42
MANU
Base Model=Idefics2, F...
2025.02
51.33
NPO
Base Model=Idefics2, F...
2025.02
50.89
KL Minim.
Base Model=Idefics2, F...
2025.02
49.64
Grad. Diff
Base Model=Idefics2, F...
2025.02
48.58
GA
Base Model=Idefics2, F...
2025.02
47.77
Feedback
Search any
task
Search any
task
