List your company

Jailbreak Defense on Llama Model Jailbreak Evaluation Prompts

3ASRPatcher
2.247.3712.517.63Jun 2, 2026

Evaluation results

MethodLinks
PatcherAttack Type=TAP2026.063
UnpatchedAttack Type=TAP2026.064
OneShotAttack Type=TAP2026.064
FTℓ∞Attack Type=TAP2026.065
MudjackingAttack Type=TAP2026.065
MENDAttack Type=TAP2026.065
PatcherAttack Type=Puzzler2026.065
FTℓ2Attack Type=TAP2026.066
ROMEAttack Type=TAP2026.066
UnpatchedAttack Type=Puzzler2026.066
OneShotAttack Type=Puzzler2026.066
PatcherAttack Type=PAIR2026.067
FTAttack Type=TAP2026.067
SPPAttack Type=TAP2026.067
MENDAttack Type=Puzzler2026.067
UnpatchedAttack Type=PAIR2026.068
FinePruningAttack Type=TAP2026.068
FTℓ∞Attack Type=Puzzler2026.068
MudjackingAttack Type=Puzzler2026.068
OneShotAttack Type=PAIR2026.069
BAERASERAttack Type=TAP2026.069
ROMEAttack Type=Puzzler2026.069
PatcherAttack Type=GCG2026.069
MENDAttack Type=PAIR2026.0610
FTℓ2Attack Type=Puzzler2026.0610
UnpatchedAttack Type=GCG2026.0610
OneShotAttack Type=GCG2026.0610
MudjackingAttack Type=PAIR2026.0611
FTAttack Type=Puzzler2026.0611
SPPAttack Type=Puzzler2026.0611
MENDAttack Type=GCG2026.0611
PatcherAttack Type=AutoDAN2026.0611
FTℓ∞Attack Type=PAIR2026.0612
ROMEAttack Type=PAIR2026.0612
FinePruningAttack Type=Puzzler2026.0612
MudjackingAttack Type=GCG2026.0612
UnpatchedAttack Type=AutoDAN2026.0612
OneShotAttack Type=AutoDAN2026.0612
FTℓ2Attack Type=PAIR2026.0613
BAERASERAttack Type=Puzzler2026.0613
FTℓ∞Attack Type=GCG2026.0613
MENDAttack Type=AutoDAN2026.0613
FTAttack Type=PAIR2026.0614
SPPAttack Type=PAIR2026.0614
ROMEAttack Type=GCG2026.0614
MudjackingAttack Type=AutoDAN2026.0614
FinePruningAttack Type=PAIR2026.0615
FTℓ2Attack Type=GCG2026.0615
FTℓ∞Attack Type=AutoDAN2026.0615
BAERASERAttack Type=PAIR2026.0616
FTAttack Type=GCG2026.0616
SPPAttack Type=GCG2026.0616
ROMEAttack Type=AutoDAN2026.0616
FinePruningAttack Type=GCG2026.0618
FTℓ2Attack Type=AutoDAN2026.0618
BAERASERAttack Type=GCG2026.0619
FTAttack Type=AutoDAN2026.0619
SPPAttack Type=AutoDAN2026.0619
FinePruningAttack Type=AutoDAN2026.0621
BAERASERAttack Type=AutoDAN2026.0622