List your company

Attack Success Rate Correlation Analysis on GCG/AutoDAN sparse-transfer diagnostic

0.965Absolute RhoTarget HRCI
-0.00740.245050.49750.74995Jun 15, 2026

Evaluation results

MethodLinks
Target HRCIRegime=R2D2, Family=geometry2026.060.965
Target fixed-source ASRRegime=R2D2, Family=behavior state2026.060.965
Target-source HRCI gapRegime=R2D2, Family=geometry gap2026.060.735
Target refusal-score driftRegime=SFT, Family=drift only2026.060.591
Target fixed-source ASRRegime=SFT, Family=behavior state2026.060.562
Target refusal-score driftRegime=R2D2, Family=drift only2026.060.482
Target-source refusal-drift gapRegime=SFT, Family=drift gap2026.060.424
Target HRCIRegime=SFT, Family=geometry2026.060.03