List your company

Explanation Faithfulness on Jailbreaking GCG (test)

0Rate of Label ChangesICL
-0.965.521218.48Mar 31, 2026

Evaluation results

MethodLinks
ICLDeletion Ratio=10%2026.030
ICLDeletion Ratio=20%2026.030
ShapleyDeletion Ratio=10%2026.0311
LIMEDeletion Ratio=10%2026.0315
EnsembleSHAPDeletion Ratio=10%2026.0315
ShapleyDeletion Ratio=20%2026.0319
LIMEDeletion Ratio=20%2026.0323
EnsembleSHAPDeletion Ratio=20%2026.0324