Our new X account is live! Follow @wizwand_team for updates

Reward Modeling on RewardBench (test)

0.933RWBench

J1-Llama-70B

Updated 4d ago

Evaluation Results

Method	Links
J1-Llama-70B 2026.01		0.933
RISE-Judge-32B 2026.01		0.927
CompassJudger2-32B 2026.01		0.926
RM-R1-Qwen-32B 2026.01		0.914
RRM-32B 2026.01		0.912
CompassJudger2-7B 2026.01		0.91
Think-J-32B 2026.01		0.905
CE-RM-4B 2026.01		0.9
CE-RM-4B 2026.01		0.894
CE-RM-4B 2026.01		0.89
RISE-Judge-7B 2026.01		0.882
GPT-4o 2026.01		0.867
DeepSeek-GRM-27B 2026.01		0.86
J1-Llama-8B 2026.01		0.857
Llama-3.3-70B-Instruct 2026.01		0.854
RM-R1-Qwen-7B 2026.01		0.852
RRM-7B 2026.01		0.822
CLoud-Gemma-2-27B 2026.01		0.82
TIR-Judge-Zero-8B 2026.01		0.814
CompassJudger1-32B 2026.01		0.812
TIR-Judge-Distill-8B 2026.01		0.81
Gemini-2.5-Flash 2026.01		0.807
TIR-Judge-Zero-4B 2026.01		0.773
TIR-Judge-Distill-4B 2026.01		0.766
JudgeLRM-7B 2026.01		0.752