ResearchBenchmarksLanguage-conditioned Long-horizon Robot Manipulation on CALVIN GPT-4 enriched (Split D)For AI assistantsMarkdownContent indexLanguage-conditioned Long-horizon Robot Manipulation on CALVIN GPT-4 enriched (Split D)Follow73.6Success Rate (1/5)LCB61.53664.66867.870.932May 8, 2024Evaluation resultsMethodMethodLinksSuccess Rate (1/5)Success Rate (2/5)Success Rate (3/5)Success Rate (4/5)Success Rate (5/5)Avg Trajectory LengthLCBFull Name=Latent Codes...Full Name=Latent Codes as Bridges2024.0573.650.228.5169.91.783DDADescription=Policy fro...Description=Policy from 3D Diffuser Actor2024.0565.239.120.311.76.11.42RFDescription=Own traini...Description=Own training of RoboFlamingo, Training Split=ABC Calvin split2024.05623316.48.64.60.4