Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Adversarial Defense on TOM-SB
Loading...
11.5
Fooling % (Hard)
Base Prompt
9.46
23.23
37
50.77
Apr 13, 2026
Fooling % (Hard)
Fooling % (All)
Updated 5d ago
Evaluation Results
Method
Method
Links
Fooling % (Hard)
Fooling % (All)
Base Prompt
Defender Backbone=Mini...
2026.04
11.5
36.7
ADA - Fooling
Defender Backbone=Mini...
2026.04
57.3
69.3
ADA - ToM Only
Defender Backbone=Mini...
2026.04
62.5
72.7
Feedback
Search any
task
Search any
task