Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Distractor Selection on Tool selection benchmark
Loading...
30
Certified Upper Bound (pu)
SecAlign
19.6
22.3
25
27.7
Oct 5, 2025
Certified Upper Bound (pu)
Updated 2mo ago
Evaluation Results
Method
Method
Links
Certified Upper Bound (pu)
SecAlign
Mitigation=SecAlign, S...
2025.10
30
Anomaly mon.
Mitigation=Anomaly mon...
2025.10
26
StruQ
Mitigation=StruQ, Sele...
2025.10
24
Perplexity
Mitigation=Perplexity,...
2025.10
23
Defended retr.
Mitigation=Defended re...
2025.10
22
None
Mitigation=None, Selec...
2025.10
20
Feedback
Search any
task
Search any
task