ResearchBenchmarksReward Modeling on ManiRewardBench LerobotFollow0.954Mean VOCTOPReward-0.03920.218650.47650.73435Feb 22, 2026Evaluation ResultsMethodMethodLinksMean VOCTOPRewardBackbone=Qwen3-VL-8BBackbone=Qwen3-VL-8B2026.020.954GVLBackbone=Gemini-2.5-Pr...Backbone=Gemini-2.5-Pro, Chat template enforced=true2026.020.62TOPRewardBackbone=Molmo2-8BBackbone=Molmo2-8B2026.020.595TOPRewardBackbone=Gemini-2.5-Pr...Backbone=Gemini-2.5-Pro, Chat template enforced=true2026.020.578GVLBackbone=Qwen3-VL-8BBackbone=Qwen3-VL-8B2026.020.332GVLBackbone=Molmo2-8BBackbone=Molmo2-8B2026.02-0.001