Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Transfer Attack on Safety Neuron Transfer Attack Evaluation distilled vs base
Loading...
8.6
ASR Before Distillation
NeuroStrike
0.696
2.748
4.8
6.852
Sep 15, 2025
ASR Before Distillation
ASR After Distillation
ASR w/o SN
Attack Ratio
Updated 2mo ago
Evaluation Results
Method
Method
Links
ASR Before Distillation
ASR After Distillation
ASR w/o SN
Attack Ratio
NeuroStrike
Base Model=Qwen2.5-Mat...
2025.09
8.6
76.7
83.1
0.4
NeuroStrike
Base Model=Qwen2.5-Mat...
2025.09
4.5
40.3
85
0.5
NeuroStrike
Base Model=Qwen2.5-32B...
2025.09
2.6
26.2
82.1
0.6
NeuroStrike
Base Model=Qwen2.5-14B...
2025.09
1.9
25.2
86.3
0.4
NeuroStrike
Base Model=Llama-3.1-8...
2025.09
1
39.3
86.9
0.7
Feedback
Search any
task
Search any
task