List your company

Reward Modeling on MR2Bench Video

50.7BoN ScoreMolmo2-4B RM
36.03639.84343.6547.457Apr 13, 2026

Evaluation results

MethodLinks
Molmo2-4B RMSize=4B2026.0450.777.40.55
Qwen3-VL-32BSize=32B2026.0449.977.80.592
InternVL3-78BSize=78B2026.0449.576.70.553
IXC-2.5-RewardSize=7B2026.0448.7740.491
Qwen3-VL-4BSize=4B2026.0447.975.80.538
Qwen3-VL-8BSize=8B2026.0447.777.20.563
Qwen3-VL-4B RMSize=4B2026.0447.573.80.482
Skywork-VL-RewardSize=7B2026.0446.774.40.499
R1-RewardSize=7B2026.0444.868.40.496
Qwen2.5-VL-7BSize=7B2026.0444.473.60.504
Molmo2-4BSize=4B2026.0443.271.90.485
Molmo2-8BSize=8B2026.0442.670.80.476
InternVL3-8BSize=8B2026.0440.471.60.466
LLaVA-CriticSize=7B2026.0440.2690.421
MM-RLHF-RewardSize=7B2026.0436.6670.352