List your company

Rerouting Detection on GSM8k (test)

1AccuracyRerouteGuard
0.46960.60730.7450.8827Jan 29, 2026

Evaluation results

MethodLinks
RerouteGuardAttack Scenario=Gray-box2026.0111
RerouteGuardAttack Scenario=Box-free2026.0111
RerouteGuardAttack Scenario=Average2026.0111
RerouteGuardAttack Scenario=White-box2026.010.990.99
LlamaGuardAttack Scenario=Gray-box2026.010.550.52
PPLAttack Scenario=Gray-box2026.010.530.1
PPLAttack Scenario=White-box2026.010.520.06
LlamaGuardAttack Scenario=Average2026.010.520.5
PPLAttack Scenario=Average2026.010.510.05
LlamaGuardAttack Scenario=White-box2026.010.510.51
PPLAttack Scenario=Box-free2026.010.50
PromptGuardAttack Scenario=Gray-box2026.010.50.67
PromptGuardAttack Scenario=Box-free2026.010.50.67
PromptGuardAttack Scenario=White-box2026.010.50.67
PromptGuardAttack Scenario=Average2026.010.50.67
LlamaGuardAttack Scenario=Box-free2026.010.490.46