ResearchBenchmarksRobot Manipulation Execution on Bridge V2 (Generalization)Follow75Visual FidelityOpenVLA28.240.3552.564.65Mar 27, 2025Evaluation ResultsMethodMethodLinksVisual FidelityMotion FidelitySemantic FidelityLanguage FidelityOpenVLAtrials per category=10...trials per category=10, scoring=partial credit2025.0375454075CoT-VLAtrials per category=10...trials per category=10, scoring=partial credit2025.0365605070Octotrials per category=10...trials per category=10, scoring=partial credit2025.033510040SUSIEtrials per category=10...trials per category=10, scoring=partial credit2025.0330102040