List your company

LLM Jailbreaking on 100-query jailbreak set

46.4Jailbreak Success RateReward-Guided RRT
25.39230.84636.341.754Jul 10, 2025

Evaluation results

MethodLinks
Reward-Guided RRTModel=DeepSeek-V32025.0746.42.74.217.7
Reward-Guided RRTModel=Gemini-2.5-Flash2025.07368.42.823.4
Baseline RRTModel=DeepSeek-V32025.0734.86.83.87.2
Reward-Guided RRTModel=Llama-3.1-70B2025.0733.813.43.427.2
Reward-Guided RRTModel=Qwen-Plus2025.07317.62.418
Baseline RRTModel=Qwen-Plus2025.0729.45.69.27.4
Baseline RRTModel=Llama-3.1-70B2025.0727.214.47.419.4
Baseline RRTModel=Gemini-2.5-Flash2025.0726.26.4514.2