Reasoning-Aware Security Evaluation
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Reasoning-Aware Security Evaluation Benign
100Processing Accuracy
11
Reasoning-Aware Security Evaluation Post-Hoc Rationalizations
0.99Det-F1
11
Reasoning-Aware Security Evaluation Fallacy Injections
100Det-F1
11
Reasoning-Aware Security Evaluation Latent Triggers
0.992Det-F1
11