List your company

Benign Compliance on WildJailbreak (WJB)

99.2Compliance RateBase
68.41676.40884.492.392Jun 9, 2025Aug 8, 2025Oct 8, 2025Dec 8, 2025Feb 6, 2026Apr 8, 2026Jun 8, 2026

Evaluation results

MethodLinks
Base2026.0699.2
Qwen2.5-3B-ITTraining=Instruction-tuned base2025.0699.2
Qwen2.5-7B-ITTraining=Instruction-tuned base2025.0699.2
Qwen2.5-14B-ITTraining=Instruction-tuned base2025.0699.2
Llama3.1-8B-IT-ABTraining=Instruction-tuned ablated base2025.0699.2
Self-RedTeamAttacker=Qwen2.5-7B2026.0698.4
Llama3.1-8B-ITTraining=Instruction-tuned base2025.0698.4
Qwen2.5-7B-IT + SELF-REDTEAMTraining=SELF-REDTEAM fine-tuned2025.0697.9
Qwen2.5-14B-IT + SELF-REDTEAMTraining=SELF-REDTEAM fine-tuned2025.0696.9
Qwen2.5-3B-IT + SELF-REDTEAMTraining=SELF-REDTEAM fine-tuned2025.0696.6
AdvGameAttacker=Qwen2.5-7B2026.0694.4
Llama3.1-8B-IT + SELF-REDTEAMTraining=SELF-REDTEAM fine-tuned2025.0693.6
Llama3.1-8B-IT-AB + SELF-REDTEAMTraining=SELF-REDTEAM fine-tuned2025.0680.6
AdvGRPO (ST)Attacker=Qwen2.5-7B2026.0672.8
AdvGRPO (ST)Attacker=Qwen2.5-14B2026.0669.6