List your company

Reward Modeling on Blessing

99.4Agreement RateAC-GenRM
62.89672.37381.8591.327Apr 7, 2026

Evaluation results

MethodLinks
AC-GenRMBackbone=Qwen3-8B2026.0499.4
AC-GenRMBackbone=Qwen3-4B2026.0499.3
AC-GenRMBackbone=Qwen3-1.7B2026.0499.1
AC-GenRMBackbone=Qwen3-0.6B2026.0498.9
DeepSeek-R1Category=Baselines2026.0497.7
GPT-4.1-MiniCategory=Baselines2026.0496.9
Claude-3.5-HaikuCategory=Baselines2026.0494.9
BaseBackbone=Qwen3-8B2026.0494.4
BaseBackbone=Qwen3-4B2026.0493.8
GPT-4.1Category=Baselines2026.0492.3
DeepSeek-V3Category=Baselines2026.0490.6
Claude-Sonnet-3.7Category=Baselines2026.0490.2
BaseBackbone=Qwen3-1.7B2026.0484.9
BaseBackbone=Qwen3-0.6B2026.0464.3