Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Priming Attack Robustness on AdvBench First-Step GCG (test)
Loading...
4
ASR (GPT-4o)
RA (ours)
2.16
14.58
27
39.42
Oct 1, 2025
ASR (GPT-4o)
ASR (Guardrail model)
ASR (Keyword matching)
Updated 4d ago
Evaluation Results
Method
Method
Links
ASR (GPT-4o)
ASR (Guardrail model)
ASR (Keyword matching)
RA (ours)
Base Model=LLaDA
2025.10
4
2
4
RA (ours)
Base Model=MMaDA
2025.10
50
46
50
Feedback
Search any
task
Search any
task