List your company

Safety Evaluation on JBench-H

0ASRDb as Alpaca
-3.218.44061.6Mar 12, 2026

Evaluation results

MethodLinks
Db as AlpacaSetting / Model=P-SFT, Llama2, ASR evaluation protocol=prefill attacks2026.030
Db as AlpacaSetting / Model=RLVR, Llama3-U2026.030
Db as Our DataSetting / Model=P-SFT, Llama2, ASR evaluation protocol=prefill attacks2026.035
Db as Our DataSetting / Model=RLVR, Llama3-U2026.035
BaselineSetting / Model=P-SFT, Llama2, ASR evaluation protocol=prefill attacks2026.0336
BaselineSetting / Model=RLVR, Llama3-U2026.0380