List your company

Jailbreak Robustness on JB-R1

98.1Evaluation Score (avg@4)Self-ReSET
42.66857.05971.4585.841May 9, 2026

Evaluation results

MethodLinks
Self-ReSETBase Model=DeepSeek-R1-Distill-Llama-8B2026.0598.1
Self-ReSETBase Model=Qwen3-8B2026.0597.4
RECAPBase Model=DeepSeek-R1-Distill-Llama-8B2026.0596.3
DAPOBase Model=DeepSeek-R1-Distill-Llama-8B2026.0594.9
Self-ReSETBase Model=DeepSeek-R1-Distill-Qwen-7B2026.0593.2
STAR-1Base Model=DeepSeek-R1-Distill-Llama-8B2026.0592
STAR-1Base Model=Qwen3-8B2026.0591.9
DAPOBase Model=Qwen3-8B2026.0591.1
RECAPBase Model=Qwen3-8B2026.0590.3
STAR-1Base Model=DeepSeek-R1-Distill-Qwen-7B2026.0585.8
RECAPBase Model=DeepSeek-R1-Distill-Qwen-7B2026.0583.7
DAPOBase Model=DeepSeek-R1-Distill-Qwen-7B2026.0581.4
SafechainBase Model=Qwen3-8B2026.0573.3
BaseBase Model=Qwen3-8B2026.0564.9
SafechainBase Model=DeepSeek-R1-Distill-Llama-8B2026.0563.7
SafechainBase Model=DeepSeek-R1-Distill-Qwen-7B2026.0558
BaseBase Model=DeepSeek-R1-Distill-Llama-8B2026.0550.6
BaseBase Model=DeepSeek-R1-Distill-Qwen-7B2026.0544.8