List your company

Reward Model Evaluation on Text-Math fraction 0.2 (test)

73Pair AccuracyArena BT
59.4862.9966.570.01Jun 24, 2026

Evaluation results

MethodLinks
Arena BTReward Model=proprietary RM2026.06730.57110.86
Arena Avg RewardReward Model=proprietary RM2026.06730.57110.86
DUALEVALReward Model=proprietary RM2026.0668.20.990.7340.703
Arena BTReward Model=Skywork-Reward-V2-Qwen3-8B2026.0666.70.4570.6240.536
Arena Avg RewardReward Model=Skywork-Reward-V2-Qwen3-8B2026.0666.10.4450.6390.503
Static 2PL2026.0664.110.5710.682
DUALEVALReward Model=Skywork-Reward-V2-Qwen3-8B2026.06600.9840.6250.692