Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Model Stealing Defense on FinQA
Loading...
33.7
Agreement (%)
Output Pert.
24.756
27.078
29.4
31.722
Jun 14, 2026
Agreement (%)
Average Score
User Impact (%)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Agreement (%)
Average Score
User Impact (%)
Output Pert.
Query budget (B)=500,...
2026.06
33.7
42.6
2.1
PRADA
Query budget (B)=500,...
2026.06
30.3
43.6
0.4
DRW
Query budget (B)=500,...
2026.06
30.1
43.1
4.8
HoneypotNet
Query budget (B)=500,...
2026.06
29.3
42.4
1.5
No Defense
Query budget (B)=500,...
2026.06
28.4
44
0
KT w/o Breadcrumb
Query budget (B)=500,...
2026.06
26.3
41.4
0
KT (Full)†
Query budget (B)=500,...
2026.06
25.7
37.8
0
KT w/o Detection
Query budget (B)=500,...
2026.06
25.1
40.3
0
Feedback
Search any
task
Search any
task