List your company

Adversarial Detection on AutoDAN

100DSRSALO
-4235077May 2, 2026

Evaluation results

MethodLinks
SALOTarget Model=Llama-3.1-8B-Instruct, Params=<20M, Grad=false2026.05100
Linear ProbeTarget Model=Qwen2.5-7B-Instruct, Params=<1M, Grad=false2026.0599.2
Linear ProbeTarget Model=Llama-3.1-8B-Instruct, Params=<1M, Grad=false2026.0598.8
SALOTarget Model=Qwen2.5-7B-Instruct, Params=<20M, Grad=false2026.0598.8
SALOTarget Model=Mistral-7B-Instruct-v0.3, Params=<20M, Grad=false2026.0598.8
Smooth LLMTarget Model=Qwen2.5-7B-Instruct, Grad=false2026.0582
GradSafeTarget Model=Llama-3.1-8B-Instruct, Grad=true2026.0576
Smooth LLMTarget Model=Llama-3.1-8B-Instruct, Grad=false2026.0572
PPL FilterTarget Model=Llama-3.1-8B-Instruct, Grad=false2026.0567.2
No Defense (1-ASR)Target Model=Qwen2.5-7B-Instruct2026.0555.6
No Defense (1-ASR)Target Model=Llama-3.1-8B-Instruct2026.0554.4
PPL FilterTarget Model=Qwen2.5-7B-Instruct, Grad=false2026.0532.8
Linear ProbeTarget Model=Mistral-7B-Instruct-v0.3, Params=<1M, Grad=false2026.0532
GradSafeTarget Model=Qwen2.5-7B-Instruct, Grad=true2026.0510
No Defense (1-ASR)Target Model=Mistral-7B-Instruct-v0.32026.053.2
GradSafeTarget Model=Mistral-7B-Instruct-v0.3, Grad=true2026.052
Smooth LLMTarget Model=Mistral-7B-Instruct-v0.3, Grad=false2026.051.6
PPL FilterTarget Model=Mistral-7B-Instruct-v0.3, Grad=false2026.050