{"name":"Embodied First Exam","short":"EFE","tagline":"Code agents have passed many exams in the digital world. What is their first exam to enter the physical world?","summary":"The first exam for code agents in the physical world. Each question is a frozen simulation scene with a goal written for a person; the answer is a program, marked by replaying it in a fresh simulator. The questions come from many benchmark suites and are sat in two modes: privileged, with the simulator open to the agent, and standard, with only the robot.","build":{"public":true,"at":"2026-10-18T09:34:52+00:00"},"task_set":{"commit":"765fc3ce16d3d6a2da39b9aba9c1bf0812853d0b","date":"2026-10-17T19:07:38-05:00"},"totals":{"suites":29,"instances":1496,"task_dirs":2522,"robots":40,"simulators":20,"graded_trials":2190},"suites":[{"id":"behavior-1k","name":"BEHAVIOR-1K","code":"B1K","origin":"adapted","example":false,"blurb":"Full household activities in house-scale scenes, for a mobile two-armed robot.","description":"100 full-length household activities in house-scale scenes. Agents must combine high-level reasoning, long-horizon navigation, and dexterous bimanual manipulation. Scored by average task success with BDDL partial credit.","edition":"2026 BEHAVIOR Challenge","simulator":"OmniGibson (NVIDIA Isaac Sim)","links":{"challenge":"https://behavior.stanford.edu/challenge/index.html","task_gallery":"https://behavior.stanford.edu/challenge/tasks/index.html","dataset_docs":"https://behavior.stanford.edu/challenge/dataset.html","call_for_participation":"https://behavior.stanford.edu/challenge/call_for_participation.html","leaderboard":"https://huggingface.co/spaces/behavior-1k/2026-challenge-leaderboard","submission":"https://behavior-1k-2026-challenge-leaderboard.hf.space/submit","repository":"https://github.com/StanfordVL/BEHAVIOR-1K","paper":"https://arxiv.org/abs/2403.09227","project_page":"https://behavior.stanford.edu/"},"robots":[{"id":"r1pro","name":"Galaxea R1 Pro","body":"mobile"}],"body":"mobile","bodies":{"mobile":100},"instances":100,"task_dirs":200,"modes":{"privileged":100,"standard":100},"oracle":{"full":0,"partial":2,"none":98},"protocols":["eai-standard/2.2"],"budgets_min":[120],"tags":{"perception-understanding":64,"planning-reasoning":97,"control-coordination":100,"feedback-adaptation":54,"mobile-whole-body-manipulation":100,"navigation-exploration":37},"on_site":true,"preview":[{"task":"setting_the_table","title":"Setting The Table","cover":true,"poster":"media/behavior-1k/setting_the_table.jpg?v=16726371","video":"media/behavior-1k/setting_the_table.mp4?v=b870e035","motion":true,"aspect":1.778,"seconds":30},{"task":"cook_brussels_sprouts","title":"Cook Brussels Sprouts","poster":"media/behavior-1k/cook_brussels_sprouts.jpg?v=3817af1a","video":"media/behavior-1k/cook_brussels_sprouts.mp4?v=b70892e7","motion":true,"aspect":1.778,"seconds":30},{"task":"make_pizza","title":"Make Pizza","poster":"media/behavior-1k/make_pizza.jpg?v=a1bd371b","video":"media/behavior-1k/make_pizza.mp4?v=8dd63967","motion":true,"aspect":1.778,"seconds":30},{"task":"slicing_vegetables","title":"Slicing Vegetables","poster":"media/behavior-1k/slicing_vegetables.jpg?v=187c0a24","video":"media/behavior-1k/slicing_vegetables.mp4?v=8cf5faa0","motion":true,"aspect":1.778,"seconds":30},{"task":"canning_food","title":"Canning Food","poster":"media/behavior-1k/canning_food.jpg?v=234f33e4","video":"media/behavior-1k/canning_food.mp4?v=906929e7","motion":true,"aspect":1.778,"seconds":30},{"task":"setup_a_bar_for_a_cocktail_party","title":"Setup A Bar For A Cocktail Party","poster":"media/behavior-1k/setup_a_bar_for_a_cocktail_party.jpg?v=f237b5c7","video":"media/behavior-1k/setup_a_bar_for_a_cocktail_party.mp4?v=8400624d","motion":true,"aspect":1.778,"seconds":30},{"task":"putting_up_Christmas_decorations_inside","title":"Putting Up Christmas Decorations Inside","poster":"media/behavior-1k/putting_up_Christmas_decorations_inside.jpg?v=b543c98e","video":"media/behavior-1k/putting_up_Christmas_decorations_inside.mp4?v=837ebaa7","motion":true,"aspect":1.778,"seconds":30},{"task":"assembling_gift_baskets","title":"Assembling Gift Baskets","poster":"media/behavior-1k/assembling_gift_baskets.jpg?v=0f0ce2c1","video":"media/behavior-1k/assembling_gift_baskets.mp4?v=b202fb38","motion":true,"aspect":1.778,"seconds":30}],"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":49,"success":8,"trials":49},"standard":{"done":49,"success":2,"trials":49}},"codex-gpt6_astra-high":{"privileged":{"done":8,"success":8,"trials":11},"standard":{"done":11,"success":10,"trials":11}},"codex-gpt6_sol-high":{"privileged":{"done":7,"success":6,"trials":11},"standard":{"done":11,"success":8,"trials":11}},"codex-gpt6_1_sol-high":{"privileged":{"done":7,"success":7,"trials":11},"standard":{"done":11,"success":10,"trials":11}}},"solved":{"privileged":11,"standard":11},"attempted":{"privileged":49,"standard":49},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":91.2,"est_n":98,"agent_s":1288862,"agent_n":98},"codex-gpt6_astra-high":{"est_usd":321.04,"est_n":19,"agent_s":61210,"agent_n":19},"codex-gpt6_sol-high":{"est_usd":160.79,"est_n":18,"agent_s":151282,"agent_n":18},"codex-gpt6_1_sol-high":{"est_usd":62.26,"est_n":18,"agent_s":87068,"agent_n":18}}},{"id":"robolab","name":"RoboLab","code":"RLB","origin":"adapted","example":false,"blurb":"Tabletop tasks that test language grounding, from colours and counts to relations and order.","description":"120 tabletop manipulation tasks for a fixed-base arm, built to test language grounding and goal specification rather than locomotion: colour and size references, counting, spatial relations, sorting, ordered stacking, reorientation. Success is a composable predicate over object state, with partial credit from scored subtasks.","edition":"v0.3.1 (2026-08-11)","simulator":"Isaac Lab 2.3.2 / Isaac Sim 5.1 (NVIDIA)","links":{"paper":"https://arxiv.org/abs/2604.09860","project_page":"https://research.nvidia.com/labs/srl/projects/robolab","leaderboard":"https://research.nvidia.com/labs/srl/projects/robolab/leaderboard.html","repository":"https://github.com/NVlabs/RoboLab","task_table":"https://github.com/NVlabs/RoboLab/blob/main/robolab/tasks/README.md"},"robots":[{"id":"franka-robotiq-2f85","name":"Franka + Robotiq 2F-85","body":"arm"}],"body":"arm","bodies":{"arm":53},"instances":53,"task_dirs":106,"modes":{"privileged":53,"standard":53},"oracle":{"full":2,"partial":0,"none":51},"protocols":["eai-standard/2.2"],"budgets_min":[60],"tags":{"perception-understanding":38,"planning-reasoning":33,"control-coordination":53,"manipulation":53,"feedback-adaptation":1},"on_site":true,"preview":[{"task":"block_stacking_specified_order","title":"Block Stacking Specified Order","cover":true,"poster":"media/robolab/block_stacking_specified_order.jpg?v=3b9b6d5b","video":"media/robolab/block_stacking_specified_order.mp4?v=b60103da","motion":true,"aspect":1.778,"seconds":16},{"task":"put_two_mugs_on_shelf","title":"Put Two Mugs On Shelf","poster":"media/robolab/put_two_mugs_on_shelf.jpg?v=208da60d","video":"media/robolab/put_two_mugs_on_shelf.mp4?v=1f29b1b5","motion":true,"aspect":1.778,"seconds":22},{"task":"cubes_and_blocks_in_bin","title":"Cubes And Blocks In Bin","poster":"media/robolab/cubes_and_blocks_in_bin.jpg?v=a9dde171","video":"media/robolab/cubes_and_blocks_in_bin.mp4?v=2c38c338","motion":true,"aspect":1.778,"seconds":18},{"task":"stack3_rubiks_cube","title":"Stack3 Rubiks Cube","poster":"media/robolab/stack3_rubiks_cube.jpg?v=72a230f5","video":"media/robolab/stack3_rubiks_cube.mp4?v=9fef99a8","motion":true,"aspect":1.778,"seconds":9},{"task":"unstack_rubiks_cube","title":"Unstack Rubiks Cube","poster":"media/robolab/unstack_rubiks_cube.jpg?v=f9dbd182","video":"media/robolab/unstack_rubiks_cube.mp4?v=0fe79027","motion":true,"aspect":1.778,"seconds":9},{"task":"block_stacking_order_agnostic","title":"Block Stacking Order Agnostic","poster":"media/robolab/block_stacking_order_agnostic.jpg?v=572015ac","video":"media/robolab/block_stacking_order_agnostic.mp4?v=a1f0680b","motion":true,"aspect":1.778,"seconds":9},{"task":"apple_and_yogurt_in_bowl","title":"Apple And Yogurt In Bowl","poster":"media/robolab/apple_and_yogurt_in_bowl.jpg?v=5e6ca2b7","video":"media/robolab/apple_and_yogurt_in_bowl.mp4?v=e957f901","motion":true,"aspect":1.778,"seconds":9},{"task":"banana_then_rubiks_cube","title":"Banana Then Rubiks Cube","poster":"media/robolab/banana_then_rubiks_cube.jpg?v=910758be","video":"media/robolab/banana_then_rubiks_cube.mp4?v=a9bf7bfc","motion":true,"aspect":1.778,"seconds":9}],"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":53,"success":44,"trials":53},"standard":{"done":53,"success":26,"trials":53}},"codex-gpt6_astra-high":{"privileged":{"done":18,"success":16,"trials":31},"standard":{"done":30,"success":17,"trials":31}},"codex-gpt6_sol-high":{"privileged":{"done":17,"success":13,"trials":25},"standard":{"done":17,"success":8,"trials":25}},"codex-gpt6_1_sol-high":{"privileged":{"done":17,"success":16,"trials":25},"standard":{"done":17,"success":9,"trials":25}}},"solved":{"privileged":52,"standard":43},"attempted":{"privileged":53,"standard":53},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":20.42,"est_n":106,"agent_s":354262,"agent_n":106},"codex-gpt6_astra-high":{"est_usd":585.34,"est_n":48,"agent_s":102507,"agent_n":48},"codex-gpt6_sol-high":{"est_usd":134.34,"est_n":34,"agent_s":102966,"agent_n":34},"codex-gpt6_1_sol-high":{"est_usd":71.51,"est_n":34,"agent_s":97963,"agent_n":34}}},{"id":"robocasa","name":"RoboCasa","code":"RCA","origin":"adapted","example":false,"blurb":"Everyday kitchen skills for a mobile manipulator in procedurally generated kitchens.","description":"Everyday kitchen tasks for a mobile manipulator in large procedurally generated kitchens: opening and closing doors and drawers, turning knobs and faucets, pressing buttons, and picking and placing between counters, cabinets, sinks, stoves and appliances, single- and multi-stage. We keep the 29 tasks with an official demonstration at the pinned commit whose demonstration re-verifies as a reference on a frozen instance (its first episode's initial state). Each task passes or fails on the environment's own success check on the state after the last action, with no partial credit, and runs in two modes: unlimited (privileged: full simulator state, unlimited resets, one replayable trajectory handed in) and limited (standard: the robot's own four cameras, one episode, no reset).","edition":"RoboCasa v0.2 @ 756598a, our selection of 29 tasks (exam @ 8c5594a43)","simulator":"robosuite 1.5.1 / MuJoCo 3.2.6 (RoboCasa v0.2)","links":{"repository":"https://github.com/robocasa/robocasa","paper":"https://arxiv.org/abs/2406.02523","project_page":"https://robocasa.ai"},"robots":[{"id":"franka-panda-omron","name":"Franka Panda on an Omron base","body":"mobile"}],"body":"mobile","bodies":{"mobile":29},"instances":29,"task_dirs":58,"modes":{"privileged":29,"standard":29},"oracle":{"full":29,"partial":0,"none":0},"protocols":["eai-standard/2.1"],"budgets_min":[60],"tags":{"planning-reasoning":5,"control-coordination":29,"manipulation":29,"perception-understanding":1},"on_site":true,"preview":[{"task":"pre_soak_pan","title":"Multi-stage · PreSoakPan","poster":"media/robocasa/pre_soak_pan.jpg?v=cf970476","video":"media/robocasa/pre_soak_pan.mp4?v=9f9ac005","motion":true,"aspect":1.778,"seconds":19},{"task":"microwave_thawing","title":"Multi-stage · MicrowaveThawing","poster":"media/robocasa/microwave_thawing.jpg?v=5d258fee","video":"media/robocasa/microwave_thawing.mp4?v=157ec665","motion":true,"aspect":1.778,"seconds":17},{"task":"prepare_coffee","title":"Multi-stage · PrepareCoffee","poster":"media/robocasa/prepare_coffee.jpg?v=21d2c6a0","video":"media/robocasa/prepare_coffee.mp4?v=a7b52753","motion":true,"aspect":1.778,"seconds":19},{"task":"turn_off_microwave","title":"Single-stage · TurnOffMicrowave","poster":"media/robocasa/turn_off_microwave.jpg?v=9012b99f","video":"media/robocasa/turn_off_microwave.mp4?v=c2cbd58a","motion":true,"aspect":1.778,"seconds":10},{"task":"coffee_setup_mug","title":"Single-stage · CoffeeSetupMug","poster":"media/robocasa/coffee_setup_mug.jpg?v=d8ae4ef4","video":"media/robocasa/coffee_setup_mug.mp4?v=91dafe02","motion":true,"aspect":1.778,"seconds":15},{"task":"turn_on_stove","title":"Single-stage · TurnOnStove","poster":"media/robocasa/turn_on_stove.jpg?v=2e6d5913","video":"media/robocasa/turn_on_stove.mp4?v=3feffcb0","motion":true,"aspect":1.778,"seconds":7},{"task":"open_double_door","title":"Single-stage · OpenDoubleDoor","poster":"media/robocasa/open_double_door.jpg?v=99b0f4bd","video":"media/robocasa/open_double_door.mp4?v=056bc526","motion":true,"aspect":1.778,"seconds":17},{"task":"arrange_vegetables","title":"Multi-stage · ArrangeVegetables","poster":"media/robocasa/arrange_vegetables.jpg?v=fcaa2980","video":"media/robocasa/arrange_vegetables.mp4?v=3a34e78b","motion":true,"aspect":1.778,"seconds":16}],"teaser":{"poster":"media/robocasa/_teaser.jpg?v=28c4b1f2","video":"media/robocasa/_teaser.mp4?v=651d5d31","cover":true,"aspect":1.778,"seconds":36},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":29,"success":25,"trials":29},"standard":{"done":29,"success":5,"trials":29}},"codex-gpt6_astra-high":{"privileged":{"done":11,"success":11,"trials":25},"standard":{"done":25,"success":18,"trials":25}},"codex-gpt6_sol-high":{"privileged":{"done":10,"success":10,"trials":25},"standard":{"done":24,"success":9,"trials":25}},"codex-gpt6_1_sol-high":{"privileged":{"done":10,"success":10,"trials":25},"standard":{"done":25,"success":15,"trials":25}}},"solved":{"privileged":29,"standard":26},"attempted":{"privileged":29,"standard":29},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":8.61,"est_n":58,"agent_s":137648,"agent_n":58},"codex-gpt6_astra-high":{"est_usd":331.22,"est_n":36,"agent_s":55994,"agent_n":36},"codex-gpt6_sol-high":{"est_usd":87.97,"est_n":34,"agent_s":64968,"agent_n":34},"codex-gpt6_1_sol-high":{"est_usd":59.74,"est_n":35,"agent_s":76376,"agent_n":35}}},{"id":"robocasa365","name":"RoboCasa365","code":"R365","origin":"adapted","example":false,"blurb":"Atomic and long composite kitchen activities from RoboCasa's 365-task suite.","description":"RoboCasa's 365-task kitchen suite for the same mobile manipulator: atomic skills and long composite activities across cooking, cleaning, storing and serving. We keep 50 of the families with a validated demonstration, chosen by a fixed rule (12 atomic, 38 composite, in SHA-256 order of their names), one frozen instance each. Each task passes or fails on the environment's own success check on the state after the last action, with no partial credit, and runs in two modes: unlimited (privileged: full simulator state, unlimited resets, one replayable trajectory handed in) and limited (standard: the robot's own four cameras, one episode, no reset).","edition":"RoboCasa365 (robocasa 1.0.1 @ 4f8a298), our selection of 50 tasks (exam @ 8c5594a43)","simulator":"robosuite 1.5.2 / MuJoCo 3.3.1 (RoboCasa365)","links":{"repository":"https://github.com/robocasa/robocasa","project":"https://robocasa.ai","paper":"https://arxiv.org/abs/2603.04356","project_page":"https://robocasa.ai"},"robots":[{"id":"franka-panda-omron","name":"Franka Panda on an Omron base","body":"mobile"}],"body":"mobile","bodies":{"mobile":50},"instances":50,"task_dirs":101,"modes":{"privileged":50,"standard":50,"standard-realtime":1},"oracle":{"full":50,"partial":0,"none":0},"protocols":["eai-standard/2.1","eai-standard/2.3"],"budgets_min":[60],"tags":{"control-coordination":50,"manipulation":28,"planning-reasoning":38,"mobile-whole-body-manipulation":22,"perception-understanding":2},"on_site":true,"preview":[{"task":"seasoning_spice_setup","title":"Composite · SeasoningSpiceSetup","poster":"media/robocasa365/seasoning_spice_setup.jpg?v=997206ef","video":"media/robocasa365/seasoning_spice_setup.mp4?v=8a8580eb","motion":true,"aspect":1.778,"seconds":26},{"task":"searing_meat","title":"Composite · SearingMeat","poster":"media/robocasa365/searing_meat.jpg?v=1a0f78b8","video":"media/robocasa365/searing_meat.mp4?v=6e0efdf9","motion":true,"aspect":1.778,"seconds":18},{"task":"make_ice_lemonade","title":"Composite · MakeIceLemonade","poster":"media/robocasa365/make_ice_lemonade.jpg?v=08bf7f1e","video":"media/robocasa365/make_ice_lemonade.mp4?v=3ea849e8","motion":true,"aspect":1.778,"seconds":22},{"task":"reheat_meal","title":"Composite · ReheatMeal","poster":"media/robocasa365/reheat_meal.jpg?v=b6bcbbe6","video":"media/robocasa365/reheat_meal.mp4?v=57882b31","motion":true,"aspect":1.778,"seconds":17},{"task":"cereal_and_bowl","title":"Composite · CerealAndBowl","poster":"media/robocasa365/cereal_and_bowl.jpg?v=86522eb6","video":"media/robocasa365/cereal_and_bowl.mp4?v=79a9bab5","motion":true,"aspect":1.778,"seconds":22},{"task":"microwave_defrost_meat","title":"Composite · MicrowaveDefrostMeat","poster":"media/robocasa365/microwave_defrost_meat.jpg?v=ac52a5c0","video":"media/robocasa365/microwave_defrost_meat.mp4?v=45278900","motion":true,"aspect":1.778,"seconds":20},{"task":"retrieve_meat","title":"Composite · RetrieveMeat","poster":"media/robocasa365/retrieve_meat.jpg?v=b4dcfa00","video":"media/robocasa365/retrieve_meat.mp4?v=c0049e4f","motion":true,"aspect":1.778,"seconds":16},{"task":"deliver_brewed_coffee","title":"Composite · DeliverBrewedCoffee","poster":"media/robocasa365/deliver_brewed_coffee.jpg?v=dcd03c69","video":"media/robocasa365/deliver_brewed_coffee.mp4?v=10a1c2be","motion":true,"aspect":1.778,"seconds":17}],"teaser":{"poster":"media/robocasa365/_teaser.jpg?v=e969c0a2","video":"media/robocasa365/_teaser.mp4?v=1e58ecf8","cover":true,"aspect":1.778,"seconds":34},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":50,"success":25,"trials":50},"standard":{"done":50,"success":0,"trials":50}},"codex-gpt6_astra-high":{"privileged":{"done":38,"success":38,"trials":50},"standard":{"done":50,"success":27,"trials":50}},"codex-gpt6_sol-high":{"privileged":{"done":17,"success":10,"trials":26},"standard":{"done":17,"success":5,"trials":26}},"codex-gpt6_1_sol-high":{"privileged":{"done":19,"success":18,"trials":27},"standard":{"done":17,"success":7,"trials":27}}},"solved":{"privileged":50,"standard":28},"attempted":{"privileged":50,"standard":50},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":19.29,"est_n":100,"agent_s":300433,"agent_n":100},"codex-gpt6_astra-high":{"est_usd":953.85,"est_n":88,"agent_s":165276,"agent_n":88},"codex-gpt6_sol-high":{"est_usd":138.27,"est_n":34,"agent_s":94994,"agent_n":34},"codex-gpt6_1_sol-high":{"est_usd":80.89,"est_n":36,"agent_s":109979,"agent_n":36}}},{"id":"robocasa-gr1","name":"RoboCasa-GR1","code":"RGR1","origin":"adapted","example":false,"blurb":"A Fourier GR1 humanoid with dexterous hands, picking and placing at a counter.","description":"A Fourier GR1 humanoid with two dexterous hands at a kitchen counter: picking objects from one container or surface and placing them in another, including novel objects, and placing into a cabinet, drawer or microwave and closing it. We keep all 24 environments with an official demonstration, one frozen instance each (the demonstration's initial state). Each task passes or fails on the environment's own success check on the state after the last action, with no partial credit, and runs in two modes: unlimited (privileged: full simulator state, unlimited resets, one replayable trajectory handed in) and limited (standard: the humanoid's own six cameras, one episode, no reset).","edition":"RoboCasa GR1 tabletop tasks @ 4840e67, all 24 (exam @ 8c5594a43)","simulator":"robosuite @ 51cc017 (1.5.1, mink whole-body IK) / MuJoCo 3.2.6 (RoboCasa GR1 tabletop)","links":{"repository":"https://github.com/robocasa/robocasa-gr1-tabletop-tasks","paper":"https://arxiv.org/abs/2503.14734","project_page":"https://developer.nvidia.com/isaac/gr00t"},"robots":[{"id":"fourier-gr1","name":"Fourier GR1","body":"humanoid"}],"body":"humanoid","bodies":{"humanoid":24},"instances":24,"task_dirs":48,"modes":{"privileged":24,"standard":24},"oracle":{"full":24,"partial":0,"none":0},"protocols":["eai-standard/2.1"],"budgets_min":[60],"tags":{"planning-reasoning":6,"control-coordination":24,"manipulation":24},"on_site":true,"preview":[{"task":"potato_to_microwave_close","title":"GR1 · PnPPotatoToMicrowaveClose","poster":"media/robocasa-gr1/potato_to_microwave_close.jpg?v=c99675d0","video":"media/robocasa-gr1/potato_to_microwave_close.mp4?v=c0cea753","motion":true,"aspect":1.778,"seconds":13},{"task":"wine_to_cabinet_close","title":"GR1 · PnPWineToCabinetClose","poster":"media/robocasa-gr1/wine_to_cabinet_close.jpg?v=a21feabe","video":"media/robocasa-gr1/wine_to_cabinet_close.mp4?v=433fb7e8","motion":true,"aspect":1.778,"seconds":15},{"task":"plate_to_bowl","title":"GR1 · PosttrainPnPNovelFromPlateToBowlSplitA","poster":"media/robocasa-gr1/plate_to_bowl.jpg?v=5f38f223","video":"media/robocasa-gr1/plate_to_bowl.mp4?v=2d0381da","motion":true,"aspect":1.778,"seconds":9},{"task":"cup_to_drawer_close","title":"GR1 · PnPCupToDrawerClose","poster":"media/robocasa-gr1/cup_to_drawer_close.jpg?v=074ec1bc","video":"media/robocasa-gr1/cup_to_drawer_close.mp4?v=c8e0dc6b","motion":true,"aspect":1.778,"seconds":12},{"task":"tray_to_tieredshelf","title":"GR1 · PosttrainPnPNovelFromTrayToTieredshelfSplitA","poster":"media/robocasa-gr1/tray_to_tieredshelf.jpg?v=1f268848","video":"media/robocasa-gr1/tray_to_tieredshelf.mp4?v=647a3f26","motion":true,"aspect":1.778,"seconds":11},{"task":"plate_to_cardboardbox","title":"GR1 · PosttrainPnPNovelFromPlateToCardboardboxSplitA","poster":"media/robocasa-gr1/plate_to_cardboardbox.jpg?v=53e03d00","video":"media/robocasa-gr1/plate_to_cardboardbox.mp4?v=f6cb31d3","motion":true,"aspect":1.778,"seconds":11},{"task":"placemat_to_tieredshelf","title":"GR1 · PosttrainPnPNovelFromPlacematToTieredshelfSplitA","poster":"media/robocasa-gr1/placemat_to_tieredshelf.jpg?v=a0b59892","video":"media/robocasa-gr1/placemat_to_tieredshelf.mp4?v=6aa590e5","motion":true,"aspect":1.778,"seconds":10},{"task":"bottle_to_cabinet_close","title":"GR1 · PnPBottleToCabinetClose","poster":"media/robocasa-gr1/bottle_to_cabinet_close.jpg?v=8af0dc71","video":"media/robocasa-gr1/bottle_to_cabinet_close.mp4?v=b5b68cd5","motion":true,"aspect":1.778,"seconds":14}],"teaser":{"poster":"media/robocasa-gr1/_teaser.jpg?v=c99675d0","video":"media/robocasa-gr1/_teaser.mp4?v=c0cea753","cover":true,"aspect":1.778,"seconds":13},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":24,"success":13,"trials":24},"standard":{"done":24,"success":1,"trials":24}},"codex-gpt6_astra-high":{"privileged":{"done":16,"success":14,"trials":24},"standard":{"done":23,"success":11,"trials":24}},"codex-gpt6_sol-high":{"privileged":{"done":15,"success":15,"trials":20},"standard":{"done":17,"success":2,"trials":20}},"codex-gpt6_1_sol-high":{"privileged":{"done":15,"success":15,"trials":24},"standard":{"done":23,"success":9,"trials":24}}},"solved":{"privileged":24,"standard":16},"attempted":{"privileged":24,"standard":24},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":9.42,"est_n":48,"agent_s":150000,"agent_n":48},"codex-gpt6_astra-high":{"est_usd":430.76,"est_n":39,"agent_s":98552,"agent_n":39},"codex-gpt6_sol-high":{"est_usd":69.57,"est_n":32,"agent_s":70089,"agent_n":32},"codex-gpt6_1_sol-high":{"est_usd":63.55,"est_n":38,"agent_s":110983,"agent_n":38}}},{"id":"metaworldplus","name":"MetaWorld+","code":"MW+","origin":"ours","example":false,"blurb":"All 50 Meta-World v3 environments for a Sawyer arm, one frozen instance each.","description":"All 50 of Meta-World's v3 environments for a Sawyer arm (reward v2), one frozen instance each: push, pick and place, open doors, drawers and windows, press buttons, turn dials and faucets, insert pegs, hammer, sweep and more. Each task passes or fails on Meta-World's own success check on the state after the agent's last action, with no partial credit, and runs in two modes: unlimited (privileged: full simulator state, unlimited resets, one replayable trajectory of at most 500 actions handed in) and limited (standard: a robot behind a service, four fixed cameras and the arm's own state, one episode, no reset). Privileged is reported as a control; the standard mode is the benchmark.","edition":"our benchmark (exam @ 8c5594a43): Meta-World v3 @ 59fc34d, all 50 environments","simulator":"MuJoCo 3.3.0 (Meta-World v3)","links":{"repository":"https://github.com/Farama-Foundation/Metaworld","paper":"https://arxiv.org/abs/2505.11289","project_page":"https://metaworld.farama.org/"},"robots":[{"id":"sawyer","name":"Rethink Sawyer","body":"arm"}],"body":"arm","bodies":{"arm":50},"instances":50,"task_dirs":100,"modes":{"privileged":50,"standard":50},"oracle":{"full":50,"partial":0,"none":0},"protocols":["eai-standard/2.1"],"budgets_min":[30],"tags":{"control-coordination":50,"manipulation":50,"planning-reasoning":8},"on_site":true,"preview":[{"task":"disassemble","title":"Meta-World · Disassemble","poster":"media/metaworldplus/disassemble.jpg?v=7e71d58a","video":"media/metaworldplus/disassemble.mp4?v=47daedf0","motion":true,"aspect":1.778,"seconds":7},{"task":"basketball","title":"Meta-World · Basketball","poster":"media/metaworldplus/basketball.jpg?v=c0eb7f60","video":"media/metaworldplus/basketball.mp4?v=2f232679","motion":true,"aspect":1.778,"seconds":5},{"task":"stick_pull","title":"Meta-World · Stick pull","poster":"media/metaworldplus/stick_pull.jpg?v=ec7e8d37","video":"media/metaworldplus/stick_pull.mp4?v=9674d7ae","motion":true,"aspect":1.778,"seconds":6},{"task":"peg_insert_side","title":"Meta-World · Peg insert side","poster":"media/metaworldplus/peg_insert_side.jpg?v=38065c4d","video":"media/metaworldplus/peg_insert_side.mp4?v=f2e39651","motion":true,"aspect":1.778,"seconds":7},{"task":"handle_pull_side","title":"Meta-World · Handle pull side","poster":"media/metaworldplus/handle_pull_side.jpg?v=59d841fc","video":"media/metaworldplus/handle_pull_side.mp4?v=18da5718","motion":true,"aspect":1.778,"seconds":7},{"task":"stick_push","title":"Meta-World · Stick push","poster":"media/metaworldplus/stick_push.jpg?v=547e3996","video":"media/metaworldplus/stick_push.mp4?v=3f16701c","motion":true,"aspect":1.778,"seconds":7},{"task":"hammer","title":"Meta-World · Hammer","poster":"media/metaworldplus/hammer.jpg?v=ce41e5ac","video":"media/metaworldplus/hammer.mp4?v=c142923e","motion":true,"aspect":1.778,"seconds":7},{"task":"assembly","title":"Meta-World · Assembly","poster":"media/metaworldplus/assembly.jpg?v=69e36283","video":"media/metaworldplus/assembly.mp4?v=1a6f612c","motion":true,"aspect":1.778,"seconds":5}],"teaser":{"poster":"media/metaworldplus/_teaser.jpg?v=b12d0efe","video":"media/metaworldplus/_teaser.mp4?v=6d0ff207","cover":true,"aspect":1.778,"seconds":15},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":50,"success":50,"trials":50},"standard":{"done":50,"success":31,"trials":50}},"codex-gpt6_astra-high":{"privileged":{"done":15,"success":15,"trials":27},"standard":{"done":25,"success":23,"trials":27}},"codex-gpt6_sol-high":{"privileged":{"done":9,"success":9,"trials":23},"standard":{"done":22,"success":19,"trials":23}},"codex-gpt6_1_sol-high":{"privileged":{"done":9,"success":9,"trials":23},"standard":{"done":22,"success":19,"trials":23}}},"solved":{"privileged":50,"standard":49},"attempted":{"privileged":50,"standard":50},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":6.82,"est_n":100,"agent_s":109069,"agent_n":100},"codex-gpt6_astra-high":{"est_usd":227.78,"est_n":40,"agent_s":53556,"agent_n":40},"codex-gpt6_sol-high":{"est_usd":28.29,"est_n":31,"agent_s":22607,"agent_n":31},"codex-gpt6_1_sol-high":{"est_usd":33.67,"est_n":31,"agent_s":43494,"agent_n":31}}},{"id":"vlabench","name":"VLABench","code":"VLB","origin":"adapted","example":false,"blurb":"Manipulation from language: attributes, common sense and multi-step chores.","description":"Language-conditioned manipulation for a Franka Panda at a table: selecting by attribute, common sense, semantic and spatial descriptions, physical question answering by pressing buttons, and multi-step tasks (heating food, setting a table, playing poker). We keep the 36 of VLABench's native tasks whose recorded expert re-verifies on a frozen instance, one instance each. Each task passes or fails on VLABench's own success check (its termination condition), with no partial credit, and runs in two modes: unlimited (privileged: every entity's pose, VLABench's skills as planning primitives, unlimited resets, one replayable trajectory handed in) and limited (standard: four cameras and the arm's own state, one episode, no reset).","edition":"VLABench @ cf588fe, our selection of 36 tasks (exam @ 8c5594a43)","simulator":"MuJoCo 3.2.2 + dm_control 1.0.22 (VLABench)","links":{"repository":"https://github.com/OpenMOSS/VLABench","paper":"https://arxiv.org/abs/2412.18194","project_page":"https://vlabench.github.io/"},"robots":[{"id":"franka-panda","name":"Franka Panda","body":"arm"}],"body":"arm","bodies":{"arm":36},"instances":36,"task_dirs":72,"modes":{"privileged":36,"standard":36},"oracle":{"full":36,"partial":0,"none":0},"protocols":["eai-standard/2.1"],"budgets_min":[30],"tags":{"perception-understanding":31,"planning-reasoning":19,"control-coordination":36,"manipulation":36,"feedback-adaptation":1},"on_site":true,"preview":[{"task":"heat_food","title":"Manipulation · heat_food","poster":"media/vlabench/heat_food.jpg?v=56138533","video":"media/vlabench/heat_food.mp4?v=8cd69e79","motion":true,"aspect":1.778,"seconds":22},{"task":"texas_holdem","title":"Manipulation · texas_holdem","poster":"media/vlabench/texas_holdem.jpg?v=30594a6d","video":"media/vlabench/texas_holdem.mp4?v=1411e3ff","motion":true,"aspect":1.778,"seconds":19},{"task":"take_out_cool_drink","title":"Manipulation · take_out_cool_drink","poster":"media/vlabench/take_out_cool_drink.jpg?v=8d34b8e3","video":"media/vlabench/take_out_cool_drink.mp4?v=6d4b4b4c","motion":true,"aspect":1.778,"seconds":19},{"task":"hammer_nail_and_hang_picture","title":"Manipulation · hammer_nail_and_hang_picture","poster":"media/vlabench/hammer_nail_and_hang_picture.jpg?v=4e30013b","video":"media/vlabench/hammer_nail_and_hang_picture.mp4?v=8a05e1ac","motion":true,"aspect":1.778,"seconds":17},{"task":"book_rearrange","title":"Manipulation · book_rearrange","poster":"media/vlabench/book_rearrange.jpg?v=0b21ae72","video":"media/vlabench/book_rearrange.mp4?v=222897b2","motion":true,"aspect":1.778,"seconds":18},{"task":"select_unique_type_mahjong","title":"Manipulation · select_unique_type_mahjong","poster":"media/vlabench/select_unique_type_mahjong.jpg?v=8a567bd0","video":"media/vlabench/select_unique_type_mahjong.mp4?v=d5301de7","motion":true,"aspect":1.778,"seconds":13},{"task":"insert_flower","title":"Manipulation · insert_flower","poster":"media/vlabench/insert_flower.jpg?v=f8b31ce7","video":"media/vlabench/insert_flower.mp4?v=d9c16dcf","motion":true,"aspect":1.778,"seconds":18},{"task":"select_billiards_common_sense","title":"Common sense · select_billiards_common_sense","poster":"media/vlabench/select_billiards_common_sense.jpg?v=d4df171d","video":"media/vlabench/select_billiards_common_sense.mp4?v=855b4d11","motion":true,"aspect":1.778,"seconds":15}],"teaser":{"poster":"media/vlabench/_teaser.jpg?v=d0ea0e83","video":"media/vlabench/_teaser.mp4?v=0db6e8e9","cover":true,"aspect":1.778,"seconds":38},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":36,"success":33,"trials":36},"standard":{"done":36,"success":15,"trials":36}},"codex-gpt6_astra-high":{"privileged":{"done":16,"success":16,"trials":29},"standard":{"done":28,"success":26,"trials":29}},"codex-gpt6_sol-high":{"privileged":{"done":10,"success":9,"trials":21},"standard":{"done":17,"success":13,"trials":21}},"codex-gpt6_1_sol-high":{"privileged":{"done":10,"success":10,"trials":21},"standard":{"done":17,"success":15,"trials":21}}},"solved":{"privileged":36,"standard":34},"attempted":{"privileged":36,"standard":36},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":8.82,"est_n":72,"agent_s":131043,"agent_n":72},"codex-gpt6_astra-high":{"est_usd":124.67,"est_n":44,"agent_s":24437,"agent_n":44},"codex-gpt6_sol-high":{"est_usd":57.1,"est_n":27,"agent_s":38000,"agent_n":27},"codex-gpt6_1_sol-high":{"est_usd":10.67,"est_n":27,"agent_s":16224,"agent_n":27}}},{"id":"humanoidbench","name":"HumanoidBench","code":"HB","origin":"adapted","example":false,"blurb":"Whole-body locomotion and manipulation for a Unitree H1 humanoid.","description":"Whole-body tasks for a simulated Unitree H1 humanoid, bare or with two Shadow dexterous hands: locomotion (walking, stairs, hurdles, balancing, a maze, sitting) and whole-body manipulation (doors, cabinets, a bookshelf, pushing and lifting, a cube in each hand). Each episode is scored by HumanoidBench's own dense reward, summed over the episode, against its own success bar. We keep the 9 hardest of its 32 registered tasks plus one fixture, in three modes: privileged (full simulator state, unlimited resets, one replayable trajectory handed in), standard (a robot behind a service: cameras, joint readings and an IMU, 50 resets) and standard-realtime (standard on a clock that does not stop: the robot runs at 50 Hz while the agent thinks).","edition":"cb11890 (upstream), our selection of 9 tasks and a fixture (2026-10-04; 24 until then)","simulator":"MuJoCo 3 (HumanoidBench's gymnasium environments)","links":{"paper":"https://arxiv.org/abs/2403.10506","project_page":"https://humanoid-bench.github.io","repository":"https://github.com/carlosferrazza/humanoid-bench"},"robots":[{"id":"unitree-h1","name":"Unitree H1","body":"humanoid"},{"id":"unitree-h1hand","name":"Unitree H1 with hands","body":"humanoid"}],"body":"humanoid","bodies":{"humanoid":10},"instances":10,"task_dirs":30,"modes":{"privileged":10,"standard":10,"standard-realtime":10},"oracle":{"full":1,"partial":0,"none":9},"protocols":["eai-standard/2.1","eai-standard/2.3"],"budgets_min":[60],"tags":{"control-coordination":10,"feedback-adaptation":2,"locomotion-stability":4,"planning-reasoning":4,"mobile-whole-body-manipulation":6,"navigation-exploration":1},"on_site":true,"preview":[{"task":"manip_window","title":"Manipulation · Window","poster":"media/humanoidbench/manip_window.jpg?v=9c1605d2","video":"media/humanoidbench/manip_window.mp4?v=ad354392","motion":true,"aspect":1.778,"seconds":22},{"task":"loco_balance_hard","title":"Locomotion · Balance hard","poster":"media/humanoidbench/loco_balance_hard.jpg?v=97a324f4","video":"media/humanoidbench/loco_balance_hard.mp4?v=91cff7b5","motion":true,"aspect":1.778,"seconds":22},{"task":"loco_hurdle","title":"Locomotion · Hurdle","poster":"media/humanoidbench/loco_hurdle.jpg?v=8c72aaeb","video":"media/humanoidbench/loco_hurdle.mp4?v=6af25b80","motion":true,"aspect":1.778,"seconds":22},{"task":"manip_cube","title":"Manipulation · Cube","poster":"media/humanoidbench/manip_cube.jpg?v=8a494ce1","video":"media/humanoidbench/manip_cube.mp4?v=d53f7083","motion":true,"aspect":1.778,"seconds":12},{"task":"loco_walk","title":"Locomotion · Walk","poster":"media/humanoidbench/loco_walk.jpg?v=5d64b400","video":"media/humanoidbench/loco_walk.mp4?v=32b27a6e","motion":true,"aspect":1.778,"seconds":22},{"task":"loco_sit_simple","title":"Locomotion · Sit simple","poster":"media/humanoidbench/loco_sit_simple.jpg?v=4f0b4983","video":"media/humanoidbench/loco_sit_simple.mp4?v=b95d29fb","motion":true,"aspect":1.778,"seconds":22},{"task":"manip_door","title":"Manipulation · Door","poster":"media/humanoidbench/manip_door.jpg?v=6634feb3","video":"media/humanoidbench/manip_door.mp4?v=76d39ce0","motion":true,"aspect":1.778,"seconds":20},{"task":"manip_bookshelf_simple","title":"Manipulation · Bookshelf simple","poster":"media/humanoidbench/manip_bookshelf_simple.jpg?v=94c3aee9","video":"media/humanoidbench/manip_bookshelf_simple.mp4?v=0254f3eb","motion":false,"aspect":1.778,"seconds":8}],"teaser":{"poster":"media/humanoidbench/_teaser.jpg?v=9c1605d2","video":"media/humanoidbench/_teaser.mp4?v=ad354392","cover":true,"aspect":1.778,"seconds":22},"results":{"codex-gpt6_luna-medium-openrouter":{"privileged":{"done":10,"success":1,"trials":10},"standard":{"done":5,"success":0,"trials":5},"standard_reset":{"done":5,"success":1,"budget":50}},"codex-gpt6_1_sol-medium":{"privileged":{"done":10,"success":3,"trials":10},"standard_reset":{"done":10,"success":1,"budget":50}},"claude_code-claude_opus5_5-medium":{"privileged":{"done":3,"success":2,"trials":3},"standard_reset":{"done":3,"success":0,"budget":50}},"codex-gpt6_sol-medium":{"privileged":{"done":1,"success":0,"trials":3},"standard_reset":{"done":3,"success":0,"budget":50}},"codex-gpt6_luna-xhigh":{"privileged":{"done":10,"success":1,"trials":10},"standard":{"done":10,"success":0,"trials":10}},"codex-gpt6_astra-high":{"privileged":{"done":9,"success":4,"trials":10},"standard":{"done":10,"success":2,"trials":10}},"codex-gpt6_sol-high":{"privileged":{"done":9,"success":3,"trials":10},"standard":{"done":6,"success":1,"trials":10}},"codex-gpt6_1_sol-high":{"privileged":{"done":9,"success":3,"trials":10},"standard":{"done":6,"success":1,"trials":10}}},"solved":{"privileged":7,"standard":2},"attempted":{"privileged":10,"standard":10},"cost":{"codex-gpt6_luna-medium-openrouter":{"est_usd":4.99,"est_n":15,"agent_s":36619,"agent_n":15},"codex-gpt6_1_sol-medium":{"est_usd":26.63,"est_n":10,"agent_s":35913,"agent_n":10},"claude_code-claude_opus5_5-medium":{"est_usd":7.1,"est_n":3,"agent_s":4625,"agent_n":3},"codex-gpt6_sol-medium":{"est_usd":4.74,"est_n":1,"agent_s":3506,"agent_n":1},"codex-gpt6_luna-xhigh":{"est_usd":2.96,"est_n":20,"agent_s":39834,"agent_n":20},"codex-gpt6_astra-high":{"est_usd":195.21,"est_n":19,"agent_s":41538,"agent_n":19},"codex-gpt6_sol-high":{"est_usd":55.03,"est_n":15,"agent_s":32321,"agent_n":15},"codex-gpt6_1_sol-high":{"est_usd":32.16,"est_n":15,"agent_s":35040,"agent_n":15}}},{"id":"molmospaces","name":"MolmoSpaces","code":"MOL","origin":"adapted","example":false,"blurb":"Zero-shot episodes in unseen houses, for a Franka FR3 and an RB-Y1 mobile robot.","description":"Ai2's zero-shot evaluation episodes for robot policies in unseen houses: a Franka FR3 picks, places (into a receptacle, next to an object, by colour), opens and closes fixtures, and a Rainbow RB-Y1 mobile manipulator navigates to an object and opens a room door. We keep one upstream episode per family, 8 in all: of three candidates a fixed rule picks per family (cluttered targets, upstream's Hard configurations, far and unseen navigation targets, three different houses), the hardest and most representative in GPT-6 Luna's runs. Each task passes or fails on upstream's own `judge_success()` at the end of the episode, with no partial credit, and runs in two modes: unlimited (privileged: the simulator's full state, upstream's kinematics, unlimited resets, one replayable trajectory handed in) and limited (standard: the robot's own cameras with depth and its proprioception, one episode, no reset).","edition":"molmo-spaces 0.2.9, benchmark molmospaces-bench-v2/20260415; our selection of 8 episodes (exam @ e46be110c,","simulator":"MuJoCo 3.5.0 (molmo-spaces 0.2.9)","links":{"repository":"https://github.com/allenai/molmospaces","paper":"https://arxiv.org/abs/2602.11337","molmobot_paper":"https://arxiv.org/abs/2603.16861","project_page":"https://allenai.github.io/molmospaces/"},"robots":[{"id":"franka-robotiq-2f85","name":"Franka + Robotiq 2F-85","body":"arm"},{"id":"rainbow-rby1","name":"Rainbow RB-Y1","body":"mobile"}],"body":"arm","bodies":{"arm":6,"mobile":2},"instances":8,"task_dirs":16,"modes":{"privileged":8,"standard":8},"oracle":{"full":0,"partial":0,"none":8},"protocols":["eai-standard/2.2"],"budgets_min":[60],"tags":{"control-coordination":8,"manipulation":6,"perception-understanding":3,"feedback-adaptation":1,"navigation-exploration":1,"mobile-whole-body-manipulation":1},"on_site":true,"preview":[{"task":"open_door","title":"RB-Y1 · Open door","poster":"media/molmospaces/open_door.jpg?v=a37f562c","video":"media/molmospaces/open_door.mp4?v=83004d5c","motion":true,"aspect":1.778,"seconds":14},{"task":"close","title":"Franka · Close","poster":"media/molmospaces/close.jpg?v=8989e962","video":"media/molmospaces/close.mp4?v=888c82ac","motion":true,"aspect":1.778,"seconds":7},{"task":"open","title":"Franka · Open","poster":"media/molmospaces/open.jpg?v=50a3ea3e","video":"media/molmospaces/open.mp4?v=5776a52e","motion":true,"aspect":1.778,"seconds":13},{"task":"pick","title":"Franka · Pick","poster":"media/molmospaces/pick.jpg?v=31349346","video":"media/molmospaces/pick.mp4?v=a7825ed5","motion":true,"aspect":1.778,"seconds":14},{"task":"pick_and_place","title":"Franka · Pick and place","poster":"media/molmospaces/pick_and_place.jpg?v=13dd0118","video":"media/molmospaces/pick_and_place.mp4?v=bb4917ec","motion":true,"aspect":1.778,"seconds":20},{"task":"pick_and_place_color","title":"Franka · Pick and place by colour","poster":"media/molmospaces/pick_and_place_color.jpg?v=42393f4c","video":"media/molmospaces/pick_and_place_color.mp4?v=da215a73","motion":true,"aspect":1.778,"seconds":10},{"task":"pick_and_place_next_to","title":"Franka · Pick and place next to","poster":"media/molmospaces/pick_and_place_next_to.jpg?v=192314e3","video":"media/molmospaces/pick_and_place_next_to.mp4?v=78fa3317","motion":true,"aspect":1.778,"seconds":26},{"task":"navigate_to","title":"RB-Y1 · Navigate to","poster":"media/molmospaces/navigate_to.jpg?v=9f30c208","video":"media/molmospaces/navigate_to.mp4?v=11ee5150","motion":true,"aspect":1.778,"seconds":16}],"teaser":{"poster":"media/molmospaces/_teaser.jpg?v=a37f562c","video":"media/molmospaces/_teaser.mp4?v=83004d5c","cover":true,"aspect":1.778,"seconds":14},"results":{"codex-gpt6_luna-medium-openrouter":{"privileged":{"done":8,"success":7,"trials":8},"standard":{"done":8,"success":0,"trials":8}}},"solved":{"privileged":7,"standard":0},"attempted":{"privileged":8,"standard":8},"cost":{"codex-gpt6_luna-medium-openrouter":{"est_usd":1.11,"est_n":16,"agent_s":14703,"agent_n":16}}},{"id":"kinder","name":"KinDER","code":"KND","origin":"adapted","example":false,"blurb":"Physical reasoning for planning and control: throw, balance, sweep, scoop, pack.","description":"Physical reasoning for robot planning and control, kept apart from perception and language: a simulated TidyBot++ mobile manipulator throws, balances, sweeps, scoops, stacks into cupboards, sorts, clears and packs. Each episode passes or fails on KinDER's own goal check (an object's centre inside a goal region, and the like), with no partial credit. We keep 5 of its 31 task families (the 4 hardest, scored, and one fixture), one variant each, and run each in two modes: unlimited (full simulator state and the goal check, unlimited resets, one replayable trajectory handed in) and limited (a robot behind a service: its base and wrist cameras, odometry and joint readings; one episode without resets).","edition":"5b2dbac (upstream), our selection of 4 tasks and a fixture (2026-10-04; 11 until then)","simulator":"MuJoCo 3.3 (KinDER's Dynamic3D environments); PyBullet for two kinematic tasks now excluded","links":{"paper":"https://arxiv.org/abs/2604.25788","repository":"https://github.com/Princeton-Robot-Planning-and-Learning/kindergarden","project_page":"https://prpl-group.com/kinder-site/"},"robots":[{"id":"tidybot-plus-plus","name":"TidyBot++","body":"mobile"}],"body":"mobile","bodies":{"mobile":5},"instances":5,"task_dirs":10,"modes":{"privileged":5,"standard":5},"oracle":{"full":1,"partial":0,"none":4},"protocols":["eai-standard/2.1"],"budgets_min":[30],"tags":{"planning-reasoning":3,"control-coordination":3,"mobile-whole-body-manipulation":3,"perception-understanding":1,"feedback-adaptation":1,"navigation-exploration":1},"on_site":true,"preview":[{"task":"constrained_cupboard","title":"Dynamic3D · Constrained cupboard","poster":"media/kinder/constrained_cupboard.jpg?v=63a002c4","video":"media/kinder/constrained_cupboard.mp4?v=7b25554b","motion":true,"aspect":1.778,"seconds":34},{"task":"dynamo","title":"Dynamic3D · Dynamo","poster":"media/kinder/dynamo.jpg?v=0af57458","video":"media/kinder/dynamo.mp4?v=d608d247","motion":true,"aspect":1.778,"seconds":20},{"task":"sweep_into_drawer","title":"Dynamic3D · Sweep into drawer","poster":"media/kinder/sweep_into_drawer.jpg?v=951a71cf","video":"media/kinder/sweep_into_drawer.mp4?v=641d0fa4","motion":true,"aspect":1.778,"seconds":32}],"teaser":{"poster":"media/kinder/_teaser.jpg?v=63a002c4","video":"media/kinder/_teaser.mp4?v=7b25554b","cover":true,"aspect":1.778,"seconds":34},"results":{"codex-gpt6_luna-medium-openrouter":{"privileged":{"done":3,"success":2,"trials":3},"standard":{"done":2,"success":0,"trials":2},"standard_reset":{"done":1,"success":1,"budget":50}},"codex-gpt6_1_sol-medium":{"privileged":{"done":3,"success":3,"trials":3},"standard_reset":{"done":3,"success":3,"budget":50}},"codex-gpt6_luna-xhigh":{"privileged":{"done":3,"success":3,"trials":3},"standard":{"done":3,"success":1,"trials":3}},"codex-gpt6_astra-high":{"privileged":{"done":1,"success":1,"trials":2},"standard":{"done":2,"success":0,"trials":2}},"codex-gpt6_sol-high":{"privileged":{"done":1,"success":1,"trials":2},"standard":{"done":2,"success":0,"trials":2}},"codex-gpt6_1_sol-high":{"privileged":{"done":1,"success":1,"trials":2},"standard":{"done":2,"success":1,"trials":2}}},"solved":{"privileged":3,"standard":2},"attempted":{"privileged":3,"standard":3},"cost":{"codex-gpt6_luna-medium-openrouter":{"est_usd":1.32,"est_n":5,"agent_s":10797,"agent_n":5},"codex-gpt6_1_sol-medium":{"est_usd":1.17,"est_n":3,"agent_s":1608,"agent_n":3},"codex-gpt6_luna-xhigh":{"est_usd":0.88,"est_n":6,"agent_s":13915,"agent_n":6},"codex-gpt6_astra-high":{"est_usd":16.34,"est_n":3,"agent_s":2958,"agent_n":3},"codex-gpt6_sol-high":{"est_usd":12.21,"est_n":3,"agent_s":6189,"agent_n":3},"codex-gpt6_1_sol-high":{"est_usd":3.1,"est_n":3,"agent_s":3329,"agent_n":3}}},{"id":"robopaint","name":"RoboPaint","code":"RPT","origin":"ours","example":false,"blurb":"A robot arm with a brush paints the picture it is shown, stroke by stroke, from calligraphy to oils.","description":"82 brush targets for a Franka Panda with a brush: Chinese calligraphy in kaishu (35) and xingshu (16), lettering (10), acrylic (10) and oil (11), each in both modes. The target has to be written or painted with strokes, the way a person does, not coloured in: writing fails when a long brush path is made of thin marks, and a painting has a path budget and its strokes must be sweeps, not zigzag, serpentine or contour-offset fills. Success needs the picture to match and the process rules to hold; work within the rules keeps a continuous score.","edition":"ours (2026-10-05)","simulator":"ManiSkill 3.0.1 / SAPIEN 3.0.3 + brush models (calligraphy, broad nib, libmypaint 1.6.1)","links":{"maniskill":"https://github.com/haosulab/ManiSkill","libmypaint":"https://github.com/mypaint/libmypaint","project_page":"https://robopaint.embodied-first-exam.ai"},"robots":[{"id":"franka-panda","name":"Franka Panda","body":"arm"}],"body":"arm","bodies":{"arm":43},"instances":43,"task_dirs":86,"modes":{"privileged":43,"standard":43},"oracle":{"full":0,"partial":43,"none":0},"protocols":["eai-standard/2.2"],"budgets_min":[60],"tags":{"perception-understanding":19,"planning-reasoning":19,"control-coordination":19,"manipulation":19},"on_site":true,"preview":[{"task":"oil_starrynight_i01","title":"Oil · Starrynight","poster":"media/robopaint/scenes/oil_starrynight_i01.jpg?v=b6973a86","video":"media/robopaint/oil_starrynight_i01.mp4?v=aaa344f4","motion":true,"aspect":1.778,"seconds":12},{"task":"oil_cafeterrace_i01","title":"Oil · Cafeterrace","poster":"media/robopaint/scenes/oil_cafeterrace_i01.jpg?v=1efb72c8","video":"media/robopaint/oil_cafeterrace_i01.mp4?v=794a1833","motion":true,"aspect":1.778,"seconds":12},{"task":"xingshu_huifenghechang","title":"Xingshu · Huifenghechang","poster":"media/robopaint/scenes/xingshu_huifenghechang.jpg?v=c0bdf032","video":"media/robopaint/xingshu_huifenghechang.mp4?v=9909170e","motion":true,"aspect":1.778,"seconds":12},{"task":"caoshu_beiquelindanshui","title":"Caoshu · Beiquelindanshui","poster":"media/robopaint/scenes/caoshu_beiquelindanshui.jpg?v=a98b424d","video":"media/robopaint/caoshu_beiquelindanshui.mp4?v=857a4778","motion":true,"aspect":1.778,"seconds":12},{"task":"acrylic_redfuji","title":"Acrylic · Redfuji","poster":"media/robopaint/scenes/acrylic_redfuji.jpg?v=655f4e39","video":"media/robopaint/acrylic_redfuji.mp4?v=f262b6ae","motion":true,"aspect":1.778,"seconds":12},{"task":"oil_irises_i01","title":"Oil · Irises","poster":"media/robopaint/scenes/oil_irises_i01.jpg?v=e45fada2","video":"media/robopaint/oil_irises_i01.mp4?v=781ea83f","motion":true,"aspect":1.778,"seconds":12},{"task":"xingshu_maolinxiuzhu","title":"Xingshu · Maolinxiuzhu","poster":"media/robopaint/scenes/xingshu_maolinxiuzhu.jpg?v=9687e8df","video":"media/robopaint/xingshu_maolinxiuzhu.mp4?v=c0d35375","motion":true,"aspect":1.778,"seconds":12},{"task":"oil_monalisa_i01","title":"Oil · Monalisa","poster":"media/robopaint/scenes/oil_monalisa_i01.jpg?v=b8fe9c58","video":"media/robopaint/oil_monalisa_i01.mp4?v=65572639","motion":true,"aspect":1.778,"seconds":12}],"teaser":{"poster":"media/robopaint/_teaser.jpg?v=d6bce3fa","video":"media/robopaint/_teaser.mp4?v=bac67b2c","cover":true,"aspect":1.778,"seconds":26},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":19,"success":2,"trials":19},"standard":{"done":19,"success":0,"trials":19}},"codex-gpt6_astra-high":{"privileged":{"done":19,"success":19,"trials":19},"standard":{"done":19,"success":7,"trials":19}}},"solved":{"privileged":19,"standard":7},"attempted":{"privileged":19,"standard":19},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":10.5,"est_n":38,"agent_s":147375,"agent_n":38},"codex-gpt6_astra-high":{"est_usd":319.8,"est_n":38,"agent_s":66250,"agent_n":38}}},{"id":"robotwin-2","name":"RoboTwin 2.0","code":"RTW","origin":"adapted","example":false,"blurb":"Fifty bimanual tabletop tasks for a dual-arm ALOHA-AgileX.","description":"Fifty bimanual tabletop manipulation tasks on a dual-arm ALOHA-AgileX (four other embodiments supported), each defined as one Python class: a scene builder, a success predicate and a scripted expert that RoboTwin uses to generate its public 100k-episode dataset. Domain randomisation (clutter, textures, lighting, table height) is a switch we leave off. Success is the task's own `check_success()` at the end of the episode, all-or-nothing.","edition":"ICML 2026 paper (arXiv 2506.18088); code `main` @ 6dde571 (2026-09-14); assets HF revision 981c92a","simulator":"SAPIEN 3 (PhysX) + CuRobo","links":{"paper":"https://arxiv.org/abs/2506.18088","project_page":"https://robotwin-platform.github.io/","documentation":"https://robotwin-platform.github.io/doc/","tasks_doc":"https://robotwin-platform.github.io/doc/tasks/index.html","repository":"https://github.com/RoboTwin-Platform/RoboTwin","dataset":"https://huggingface.co/datasets/TianxingChen/RoboTwin2.0","leaderboard":"https://robotwin-platform.github.io/leaderboard"},"robots":[{"id":"aloha-agilex","name":"ALOHA AgileX","body":"bimanual"}],"body":"bimanual","bodies":{"bimanual":47},"instances":47,"task_dirs":94,"modes":{"privileged":47,"standard":47},"oracle":{"full":47,"partial":0,"none":0},"protocols":["eai-standard/2.1"],"budgets_min":[30],"tags":{"control-coordination":47,"manipulation":47,"planning-reasoning":21,"perception-understanding":18,"interaction-collaboration":2,"feedback-adaptation":1},"on_site":true,"preview":[{"task":"hanging_mug","title":"Hanging mug","poster":"media/robotwin-2/hanging_mug.jpg?v=3f4d4f15","video":"media/robotwin-2/hanging_mug.mp4?v=1c902f95","motion":true,"aspect":1.778,"seconds":39},{"task":"put_bottles_dustbin","title":"Put bottles dustbin","poster":"media/robotwin-2/put_bottles_dustbin.jpg?v=5fe48754","video":"media/robotwin-2/put_bottles_dustbin.mp4?v=c2b27f1a","motion":true,"aspect":1.778,"seconds":40},{"task":"scan_object","title":"Scan object","poster":"media/robotwin-2/scan_object.jpg?v=545f5e6a","video":"media/robotwin-2/scan_object.mp4?v=c0cd3a04","motion":true,"aspect":1.778,"seconds":46},{"task":"dump_bin_bigbin","title":"Dump bin big bin","poster":"media/robotwin-2/dump_bin_bigbin.jpg?v=10d420fb","video":"media/robotwin-2/dump_bin_bigbin.mp4?v=fcd4648a","motion":true,"aspect":1.778,"seconds":59},{"task":"place_can_basket","title":"Place can basket","poster":"media/robotwin-2/place_can_basket.jpg?v=abae6e3e","video":"media/robotwin-2/place_can_basket.mp4?v=db4868ba","motion":true,"aspect":1.778,"seconds":35},{"task":"stack_blocks_three","title":"Stack blocks three","poster":"media/robotwin-2/stack_blocks_three.jpg?v=cfcc5941","video":"media/robotwin-2/stack_blocks_three.mp4?v=f40f5f36","motion":true,"aspect":1.778,"seconds":25},{"task":"blocks_ranking_rgb","title":"Blocks ranking rgb","poster":"media/robotwin-2/blocks_ranking_rgb.jpg?v=dbdc8c64","video":"media/robotwin-2/blocks_ranking_rgb.mp4?v=d2d4d8ce","motion":true,"aspect":1.778,"seconds":21},{"task":"pick_diverse_bottles","title":"Pick diverse bottles","poster":"media/robotwin-2/pick_diverse_bottles.jpg?v=efbe723e","video":"media/robotwin-2/pick_diverse_bottles.mp4?v=c1448319","motion":true,"aspect":1.778,"seconds":36}],"teaser":{"poster":"media/robotwin-2/_teaser.jpg?v=3f4d4f15","video":"media/robotwin-2/_teaser.mp4?v=1c902f95","cover":true,"aspect":1.778,"seconds":39},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":45,"success":40,"trials":45},"standard":{"done":45,"success":16,"trials":45}},"codex-gpt6_luna-medium":{"privileged":{"done":21,"success":11,"trials":22},"standard":{"done":22,"success":3,"trials":22}},"codex-gpt6_luna-xhigh-azure":{"privileged":{"done":47,"success":42,"trials":47},"standard":{"done":47,"success":20,"trials":47}},"codex-gpt6_astra-high":{"privileged":{"done":12,"success":12,"trials":31},"standard":{"done":30,"success":23,"trials":31}},"codex-gpt6_sol-high":{"privileged":{"done":11,"success":11,"trials":21},"standard":{"done":17,"success":3,"trials":21}},"codex-gpt6_1_sol-high":{"privileged":{"done":11,"success":11,"trials":21},"standard":{"done":17,"success":12,"trials":21}}},"solved":{"privileged":47,"standard":43},"attempted":{"privileged":47,"standard":47},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":10.51,"est_n":90,"agent_s":160129,"agent_n":90},"codex-gpt6_luna-medium":{"est_usd":1.9,"est_n":43,"agent_s":32168,"agent_n":43},"codex-gpt6_luna-xhigh-azure":{"est_usd":8.0,"est_n":94,"agent_s":130256,"agent_n":94},"codex-gpt6_astra-high":{"est_usd":218.51,"est_n":42,"agent_s":36716,"agent_n":42},"codex-gpt6_sol-high":{"est_usd":76.47,"est_n":28,"agent_s":50447,"agent_n":28},"codex-gpt6_1_sol-high":{"est_usd":17.44,"est_n":28,"agent_s":21464,"agent_n":28}}},{"id":"mujoco-playground","name":"MuJoCo Playground","code":"MJP","origin":"adapted","example":false,"blurb":"Quadrupeds that walk, get up and balance, and manipulation for ALOHA 2 and a Franka Panda.","description":"Thirteen environments of Google DeepMind's MuJoCo Playground, run in plain MuJoCo on the CPU. Nine are quadruped locomotion: the Unitree Go1 tracks a joystick on flat and rough ground, gets up from a fall and balances in a handstand and a footstand; Boston Dynamics Spot tracks a joystick, with and without a commanded gait, and gets up; Google's Barkour vB tracks a joystick while it is pushed. Four are manipulation: ALOHA 2 hand-over and peg insertion, and a Franka Panda that opens a cabinet and places a cube in a given orientation. Playground trains with dense rewards, so each task has a success test of ours, built from its termination rules and reward targets.","edition":"MuJoCo Playground (arXiv 2502.08844); code `main` @ 4057c14","simulator":"MuJoCo 3.3.7 (plain MuJoCo on the CPU; upstream trains in MJX / MuJoCo Warp)","links":{"paper":"https://arxiv.org/abs/2502.08844","project_page":"https://playground.mujoco.org/","repository":"https://github.com/google-deepmind/mujoco_playground"},"robots":[{"id":"aloha-2","name":"ALOHA 2","body":"bimanual"},{"id":"boston-dynamics-spot","name":"Boston Dynamics Spot","body":"quadruped"},{"id":"franka-panda","name":"Franka Panda","body":"arm"},{"id":"google-barkour-vb","name":"Google Barkour vB","body":"quadruped"},{"id":"unitree-go1","name":"Unitree Go1","body":"quadruped"}],"body":"quadruped","bodies":{"bimanual":2,"arm":2,"quadruped":9},"instances":13,"task_dirs":27,"modes":{"privileged":13,"standard":13,"standard-realtime":1},"oracle":{"full":13,"partial":0,"none":0},"protocols":["eai-standard/2.1","eai-standard/2.2","eai-standard/2.3"],"budgets_min":[30],"tags":{"planning-reasoning":2,"control-coordination":4,"manipulation":4,"interaction-collaboration":1},"on_site":true,"preview":[{"task":"go1_joystick_rough_terrain","title":"Go1 joystick rough terrain","poster":"media/mujoco-playground/go1_joystick_rough_terrain.jpg?v=434cc7a1","video":"media/mujoco-playground/go1_joystick_rough_terrain.mp4?v=acd56b9e","motion":true,"aspect":1.778,"seconds":22},{"task":"aloha_single_peg_insertion","title":"Aloha Single Peg Insertion","poster":"media/mujoco-playground/aloha_single_peg_insertion.jpg?v=3a1b783b","video":"media/mujoco-playground/aloha_single_peg_insertion.mp4?v=eb0825a2","motion":true,"aspect":1.778,"seconds":38},{"task":"barkour_joystick","title":"Barkour joystick","poster":"media/mujoco-playground/barkour_joystick.jpg?v=6dad717c","video":"media/mujoco-playground/barkour_joystick.mp4?v=6b79f2e5","motion":true,"aspect":1.778,"seconds":12},{"task":"spot_joystick_gait_tracking","title":"Spot joystick gait tracking","poster":"media/mujoco-playground/spot_joystick_gait_tracking.jpg?v=42630a09","video":"media/mujoco-playground/spot_joystick_gait_tracking.mp4?v=3a3ed81a","motion":true,"aspect":1.778,"seconds":22},{"task":"aloha_hand_over","title":"Aloha Hand Over","poster":"media/mujoco-playground/aloha_hand_over.jpg?v=286f8599","video":"media/mujoco-playground/aloha_hand_over.mp4?v=303288a5","motion":true,"aspect":1.778,"seconds":46},{"task":"go1_joystick_flat_terrain","title":"Go1 joystick flat terrain","poster":"media/mujoco-playground/go1_joystick_flat_terrain.jpg?v=a9194aa9","video":"media/mujoco-playground/go1_joystick_flat_terrain.mp4?v=bfd628c1","motion":true,"aspect":1.778,"seconds":22},{"task":"panda_open_cabinet","title":"Panda Open Cabinet","poster":"media/mujoco-playground/panda_open_cabinet.jpg?v=e473e67e","video":"media/mujoco-playground/panda_open_cabinet.mp4?v=ae27719b","motion":true,"aspect":1.778,"seconds":37},{"task":"panda_pick_cube_orientation","title":"Panda Pick Cube Orientation","poster":"media/mujoco-playground/panda_pick_cube_orientation.jpg?v=c9b8670f","video":"media/mujoco-playground/panda_pick_cube_orientation.mp4?v=e19596c4","motion":true,"aspect":1.778,"seconds":25}],"teaser":{"poster":"media/mujoco-playground/_teaser.jpg?v=434cc7a1","video":"media/mujoco-playground/_teaser.mp4?v=acd56b9e","cover":true,"aspect":1.778,"seconds":22},"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":4,"success":4,"trials":4},"standard":{"done":4,"success":1,"trials":4}},"codex-gpt6_luna-xhigh-azure":{"privileged":{"done":4,"success":4,"trials":4},"standard":{"done":4,"success":0,"trials":4}},"codex-gpt6_astra-high":{"privileged":{"done":4,"success":4,"trials":4},"standard":{"done":4,"success":4,"trials":4}},"codex-gpt6_sol-high":{"privileged":{"done":3,"success":3,"trials":4},"standard":{"done":4,"success":0,"trials":4}},"codex-gpt6_1_sol-high":{"privileged":{"done":3,"success":3,"trials":4},"standard":{"done":4,"success":3,"trials":4}}},"solved":{"privileged":4,"standard":4},"attempted":{"privileged":4,"standard":4},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":1.39,"est_n":8,"agent_s":19182,"agent_n":8},"codex-gpt6_luna-xhigh-azure":{"est_usd":0.7,"est_n":8,"agent_s":16696,"agent_n":8},"codex-gpt6_astra-high":{"est_usd":26.59,"est_n":8,"agent_s":5267,"agent_n":8},"codex-gpt6_sol-high":{"est_usd":15.35,"est_n":7,"agent_s":6310,"agent_n":7},"codex-gpt6_1_sol-high":{"est_usd":6.03,"est_n":7,"agent_s":7237,"agent_n":7}}},{"id":"dextoolbench","name":"DexToolBench","code":"DTB","origin":"adapted","example":false,"blurb":"Dexterous tool use with a hammer, a marker, an eraser, a brush, a spatula or a screwdriver.","description":"Twenty-four dexterous tool-use tasks from SimToolReal: 6 tool categories (hammer, marker, eraser, brush, spatula, screwdriver) x 2 real objects x 2 motions. Each task is a start pose and a sequence of goal poses for the tool, extracted from a human RGB-D demo; the score is task progress, the fraction of goals reached in order. Upstream runs it in Isaac Sim / Isaac Gym; we run SimToolReal's own MuJoCo sim2sim scene, rebuilt, with each category's table props.","edition":"SimToolReal (arXiv 2602.16863); code `main` @ 313d5ae (2026-08-13)","simulator":"MuJoCo 3.3.7 (our port; upstream runs Isaac Sim / Isaac Gym)","links":{"paper":"https://arxiv.org/abs/2602.16863","project_page":"https://simtoolreal.github.io/","repository":"https://github.com/tylerlum/simtoolreal"},"robots":[{"id":"kuka-iiwa14-sharpa-ha4","name":"KUKA iiwa14 + Sharpa HA4 hand","body":"hand"}],"body":"hand","bodies":{"hand":21},"instances":21,"task_dirs":42,"modes":{"privileged":21,"standard":21},"oracle":{"full":6,"partial":15,"none":0},"protocols":["eai-standard/2.1"],"budgets_min":[60],"tags":{"planning-reasoning":21,"control-coordination":21,"manipulation":21},"on_site":true,"preview":[{"task":"hammer_claw_hammer_swing_down","title":"Claw hammer · swing down","poster":"media/dextoolbench/hammer_claw_hammer_swing_down.jpg?v=9169e783","video":"media/dextoolbench/hammer_claw_hammer_swing_down.mp4?v=48034b58","motion":true,"aspect":1.333,"seconds":12},{"task":"brush_blue_brush_sweep_forward","title":"Blue brush · sweep forward","poster":"media/dextoolbench/brush_blue_brush_sweep_forward.jpg?v=8696dd45","video":"media/dextoolbench/brush_blue_brush_sweep_forward.mp4?v=c002b463","motion":true,"aspect":1.333,"seconds":14},{"task":"hammer_claw_hammer_swing_side","title":"Claw hammer · swing side","poster":"media/dextoolbench/hammer_claw_hammer_swing_side.jpg?v=f2a5a3ba","video":"media/dextoolbench/hammer_claw_hammer_swing_side.mp4?v=fd6cbcd2","motion":true,"aspect":1.778,"seconds":42},{"task":"spatula_flat_spatula_flip_over","title":"Flat spatula · flip over","poster":"media/dextoolbench/spatula_flat_spatula_flip_over.jpg?v=071280f1","video":"media/dextoolbench/spatula_flat_spatula_flip_over.mp4?v=c36afac0","motion":true,"aspect":1.333,"seconds":21},{"task":"spatula_flat_spatula_serve_plate","title":"Flat spatula · serve plate","poster":"media/dextoolbench/spatula_flat_spatula_serve_plate.jpg?v=9ec78b1e","video":"media/dextoolbench/spatula_flat_spatula_serve_plate.mp4?v=eb65e621","motion":true,"aspect":1.333,"seconds":20},{"task":"eraser_handle_eraser_wipe_c","title":"Handle eraser · wipe c","poster":"media/dextoolbench/eraser_handle_eraser_wipe_c.jpg?v=3ef43fc3","video":"media/dextoolbench/eraser_handle_eraser_wipe_c.mp4?v=af7430bf","motion":true,"aspect":1.333,"seconds":24},{"task":"screwdriver_long_screwdriver_spin_vertical","title":"Long screwdriver · spin vertical","poster":"media/dextoolbench/screwdriver_long_screwdriver_spin_vertical.jpg?v=5ae177fe","video":"media/dextoolbench/screwdriver_long_screwdriver_spin_vertical.mp4?v=b01fddb3","motion":true,"aspect":1.333,"seconds":22},{"task":"hammer_mallet_hammer_swing_side","title":"Mallet hammer · swing side","poster":"media/dextoolbench/hammer_mallet_hammer_swing_side.jpg?v=9eb27986","video":"media/dextoolbench/hammer_mallet_hammer_swing_side.mp4?v=d4b305c1","motion":true,"aspect":1.333,"seconds":15}],"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":21,"success":0,"trials":21},"standard":{"done":21,"success":0,"trials":21}},"codex-gpt6_luna-xhigh-azure":{"privileged":{"done":21,"success":1,"trials":21},"standard":{"done":21,"success":0,"trials":21}},"codex-gpt6_astra-high":{"privileged":{"done":21,"success":14,"trials":21},"standard":{"done":21,"success":1,"trials":21}},"codex-gpt6_sol-high":{"privileged":{"done":17,"success":5,"trials":20},"standard":{"done":17,"success":0,"trials":20}},"codex-gpt6_1_sol-high":{"privileged":{"done":17,"success":10,"trials":20},"standard":{"done":17,"success":0,"trials":20}}},"solved":{"privileged":15,"standard":1},"attempted":{"privileged":21,"standard":21},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":9.13,"est_n":42,"agent_s":149096,"agent_n":42},"codex-gpt6_luna-xhigh-azure":{"est_usd":12.66,"est_n":42,"agent_s":144531,"agent_n":42},"codex-gpt6_astra-high":{"est_usd":854.18,"est_n":42,"agent_s":115159,"agent_n":42},"codex-gpt6_sol-high":{"est_usd":179.56,"est_n":34,"agent_s":107019,"agent_n":34},"codex-gpt6_1_sol-high":{"est_usd":88.76,"est_n":34,"agent_s":98454,"agent_n":34}}},{"id":"dexjoco","name":"DexJoCo","code":"DXJ","origin":"adapted","example":false,"blurb":"A 16-DoF Allegro hand on a Franka arm, one hand or two.","description":"The 11 tasks of DexJoCo for a 16-joint Allegro hand on a Franka Panda arm, six with one arm and five with two side by side: spray a plant, click a mouse, drive a nail, lift a bucket, pinch with tongs, fold glasses into their case, move a three-disk Hanoi tower, cook a hot dog in a microwave, unlock a tablet, fit a peg into a socket and take a photograph. Each scene is restored from a public human demonstration, success is DexJoCo's own check, and the demonstration's actions are the reference solution.","simulator":"MuJoCo 3.4 (DexJoCo's gymnasium environments, CPU)","links":{"repository":"https://github.com/brave-eai/dexjoco","paper":"https://arxiv.org/abs/2605.16257","project_page":"https://dexjoco.github.io/"},"robots":[{"id":"franka-allegro","name":"Franka + Allegro hand","body":"hand"},{"id":"franka-allegro-bimanual","name":"Two Franka + Allegro hands","body":"hand"}],"body":"hand","bodies":{"hand":11},"instances":11,"task_dirs":22,"modes":{"privileged":11,"standard":11},"oracle":{"full":11,"partial":0,"none":0},"protocols":["eai-standard/2.2"],"budgets_min":[60],"on_site":false,"preview":[{"task":"water_plant","title":"Water plant","poster":"media/dexjoco/water_plant.jpg?v=389861dd","video":"media/dexjoco/water_plant.mp4?v=bddb29b3","motion":true,"aspect":1.778,"seconds":8},{"task":"fold_glasses","title":"Fold glasses","poster":"media/dexjoco/fold_glasses.jpg?v=fdd9b66e","video":"media/dexjoco/fold_glasses.mp4?v=536f8a81","motion":true,"aspect":1.778,"seconds":14},{"task":"bimanual_photograph","title":"Bimanual photograph","poster":"media/dexjoco/bimanual_photograph.jpg?v=e613a880","video":"media/dexjoco/bimanual_photograph.mp4?v=fb1090a4","motion":true,"aspect":1.778,"seconds":7},{"task":"pick_bucket","title":"Pick bucket","poster":"media/dexjoco/pick_bucket.jpg?v=8ff0245e","video":"media/dexjoco/pick_bucket.mp4?v=8d2acfc3","motion":true,"aspect":1.778,"seconds":14},{"task":"hammer_nail","title":"Hammer nail","poster":"media/dexjoco/hammer_nail.jpg?v=d67060ec","video":"media/dexjoco/hammer_nail.mp4?v=9138dc97","motion":true,"aspect":1.778,"seconds":6},{"task":"bimanual_assembly","title":"Bimanual assembly","poster":"media/dexjoco/bimanual_assembly.jpg?v=1dbb7652","video":"media/dexjoco/bimanual_assembly.mp4?v=4d25802a","motion":true,"aspect":1.778,"seconds":14},{"task":"bimanual_hanoi","title":"Bimanual hanoi","poster":"media/dexjoco/bimanual_hanoi.jpg?v=5dcf791d","video":"media/dexjoco/bimanual_hanoi.mp4?v=14eef7e7","motion":true,"aspect":1.778,"seconds":24},{"task":"bimanual_microwave_cook","title":"Bimanual microwave cook","poster":"media/dexjoco/bimanual_microwave_cook.jpg?v=28c74208","video":"media/dexjoco/bimanual_microwave_cook.mp4?v=7d46ae3c","motion":true,"aspect":1.778,"seconds":19}],"teaser":{"poster":"media/dexjoco/_teaser.jpg?v=58348f2f","video":"media/dexjoco/_teaser.mp4?v=a6792c87","cover":true,"aspect":1.778,"seconds":10},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"myosuite","name":"MyoSuite","code":"MYO","origin":"adapted","example":false,"blurb":"Musculoskeletal hands, arms and legs, driven one muscle excitation at a time.","description":"Twenty-two MyoSuite tasks for musculoskeletal models of the human body: the MyoHand (39 muscles), the MyoArm, a table-tennis player with torso and arm, an arm next to a prosthetic arm, and the MyoLegs, one of them with a prosthetic leg. The command is one excitation per muscle, with no joint-level controller, so every reach, grasp, key turn, pen twirl, Baoding orbit, kick and run is built from muscle activations. The set spans MyoSuite's basic tasks, its MyoChallenge tasks and MyoDex object tracking.","simulator":"MyoSuite 2.12.2 (MuJoCo 3.6, CPU)","links":{"repository":"https://github.com/MyoHub/myosuite","paper":"https://arxiv.org/abs/2205.13600","project_page":"https://www.myosuite.org"},"robots":[{"id":"myoarm","name":"MyoArm","body":"musculoskeletal"},{"id":"myoarm-mpl","name":"MyoArm + prosthetic arm","body":"musculoskeletal"},{"id":"myoarm-torso","name":"MyoArm with torso","body":"musculoskeletal"},{"id":"myohand","name":"MyoHand","body":"musculoskeletal"},{"id":"myoleg-osl","name":"MyoLeg + prosthetic leg","body":"musculoskeletal"},{"id":"myolegs","name":"MyoLegs","body":"musculoskeletal"}],"body":"musculoskeletal","bodies":{"musculoskeletal":22},"instances":22,"task_dirs":45,"modes":{"privileged":22,"standard":22,"standard-realtime":1},"oracle":{"full":3,"partial":0,"none":19},"protocols":["eai-standard/2.2","eai-standard/2.3"],"budgets_min":[60],"on_site":false,"preview":[{"task":"osl_run","title":"Osl run","poster":"media/myosuite/osl_run.jpg?v=a21eb94a","video":"media/myosuite/osl_run.mp4?v=aec996a1","motion":true,"aspect":1.778,"seconds":25},{"task":"baoding","title":"Baoding","poster":"media/myosuite/baoding.jpg?v=3d79e192","video":"media/myosuite/baoding.mp4?v=311d9edf","motion":true,"aspect":1.778,"seconds":8},{"task":"soccer","title":"Soccer","poster":"media/myosuite/soccer.jpg?v=41117f30","video":"media/myosuite/soccer.mp4?v=41ae28f0","motion":true,"aspect":1.778,"seconds":5},{"task":"table_tennis","title":"Table tennis","poster":"media/myosuite/table_tennis.jpg?v=d3b2c321","video":"media/myosuite/table_tennis.mp4?v=101f89f7","motion":true,"aspect":1.778,"seconds":6},{"task":"relocate","title":"Relocate","poster":"media/myosuite/relocate.jpg?v=afeb6266","video":"media/myosuite/relocate.mp4?v=34516fa1","motion":true,"aspect":1.778,"seconds":7},{"task":"arm_reach","title":"Arm reach","poster":"media/myosuite/arm_reach.jpg?v=aaeebdfe","video":"media/myosuite/arm_reach.mp4?v=06a4ab7c","motion":true,"aspect":1.778,"seconds":7},{"task":"chase_tag","title":"Chase tag","poster":"media/myosuite/chase_tag.jpg?v=f7a749d0","video":"media/myosuite/chase_tag.mp4?v=9ae6cfcd","motion":true,"aspect":1.778,"seconds":6},{"task":"dex_apple_lift","title":"Dex apple lift","poster":"media/myosuite/dex_apple_lift.jpg?v=13ea205b","video":"media/myosuite/dex_apple_lift.mp4?v=f7d6b961","motion":true,"aspect":1.778,"seconds":6}],"teaser":{"poster":"media/myosuite/_teaser.jpg?v=a21eb94a","video":"media/myosuite/_teaser.mp4?v=aec996a1","cover":true,"aspect":1.778,"seconds":25},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"roboarc","name":"RoboARC","code":"ARC","origin":"ours","example":false,"blurb":"Topological puzzles, ropes, chains, cloth and IKEA furniture: separate, thread, untie, fold, assemble.","description":"Topological manipulation puzzles (formerly TopoGym and RoboTangle): free a coil from a cast, open interlocked claws, take a ring off a horseshoe, get a hedgehog out of its cage, pick a lock, guide a loop along a buzz wire, bundle fibers, thread and untie ropes, take rings off slotted posts, and flatten or fold a blanket. With them, IKEA furniture assembly for two arms with physical grasps: screw the legs into a LACK table, fit the dowels and panels of an EKET cabinet, drive screws into a VARIERA box with a screwdriver, close the snaps of a VESKEN shelf. ALOHA 2 or Franka Panda arms, on MuJoCo with the cloth on a deterministic GPU solver.","simulator":"MuJoCo 3.14 (CPU float64; the cloth on a deterministic GPU solver); videos rendered with Blender Cycles","links":{"project_page":"https://roboarc.embodied-first-exam.ai"},"robots":[{"id":"aloha-2","name":"ALOHA 2","body":"bimanual"},{"id":"franka-robotiq-2f85","name":"Franka + Robotiq 2F-85","body":"arm"},{"id":"franka-panda-bimanual","name":"Two Franka Panda arms","body":"bimanual"}],"body":"bimanual","bodies":{"bimanual":73,"arm":33},"instances":106,"task_dirs":212,"modes":{"privileged":106,"standard":106},"oracle":{"full":96,"partial":0,"none":2},"protocols":["eai-standard/2.4"],"budgets_min":[180],"on_site":false,"preview":[{"task":"interlocked_claws_12","title":"Interlocked claws 12","poster":"media/roboarc/interlocked_claws_12.jpg?v=196ebbd3","video":"media/roboarc/interlocked_claws_12.mp4?v=7603ae7e","motion":true,"aspect":1.778,"seconds":42},{"task":"interlocked_claws_10","title":"Interlocked claws 10","poster":"media/roboarc/interlocked_claws_10.jpg?v=e1a21011","video":"media/roboarc/interlocked_claws_10.mp4?v=e7f834d1","motion":true,"aspect":1.778,"seconds":22},{"task":"cast_keyhole_06","title":"Cast keyhole 06","poster":"media/roboarc/cast_keyhole_06.jpg?v=94d07429","video":"media/roboarc/cast_keyhole_06.mp4?v=0003857a","motion":true,"aspect":1.778,"seconds":28},{"task":"rope_threading_08","title":"Rope threading 08","poster":"media/roboarc/rope_threading_08.jpg?v=05d24f3c","video":"media/roboarc/rope_threading_08.mp4?v=a944dae5","motion":true,"aspect":1.778,"seconds":33},{"task":"knot_untying_05","title":"Knot untying 05","poster":"media/roboarc/knot_untying_05.jpg?v=57549292","video":"media/roboarc/knot_untying_05.mp4?v=212734c8","motion":true,"aspect":1.778,"seconds":17},{"task":"hedgehog_in_cage_05","title":"Hedgehog in cage 05","poster":"media/roboarc/hedgehog_in_cage_05.jpg?v=a669931c","video":"media/roboarc/hedgehog_in_cage_05.mp4?v=24fdacee","motion":true,"aspect":1.778,"seconds":22},{"task":"slotted_post_ring_05","title":"Slotted post ring 05","poster":"media/roboarc/slotted_post_ring_05.jpg?v=5d19b73f","video":"media/roboarc/slotted_post_ring_05.mp4?v=72d8c278","motion":true,"aspect":1.778,"seconds":32},{"task":"buzz_wire_3d_10","title":"Buzz wire 3d 10","poster":"media/roboarc/buzz_wire_3d_10.jpg?v=b724eb83","video":"media/roboarc/buzz_wire_3d_10.mp4?v=1e3f1db1","motion":true,"aspect":1.778,"seconds":42},{"task":"vesken","title":"Vesken","poster":"media/ikea-mujoco/vesken.jpg?v=d182d0ac","video":"media/ikea-mujoco/vesken.mp4?v=90eafbc3","motion":true,"aspect":1.778,"seconds":42},{"task":"gladom","title":"Gladom","poster":"media/ikea-mujoco/gladom.jpg?v=97a07d79","video":"media/ikea-mujoco/gladom.mp4?v=f1101bbd","motion":true,"aspect":1.778,"seconds":22},{"task":"eket","title":"Eket","poster":"media/ikea-mujoco/eket.jpg?v=5d524e83","video":"media/ikea-mujoco/eket.mp4?v=fe308d41","motion":true,"aspect":1.778,"seconds":24},{"task":"ikea365","title":"Ikea365","poster":"media/ikea-mujoco/ikea365.jpg?v=db65d45a","video":"media/ikea-mujoco/ikea365.mp4?v=889ba323","motion":true,"aspect":1.778,"seconds":30},{"task":"variera","title":"Variera","poster":"media/ikea-mujoco/variera.jpg?v=6d4c6343","video":"media/ikea-mujoco/variera.mp4?v=de422434","motion":true,"aspect":1.778,"seconds":22},{"task":"stegesund","title":"Stegesund","poster":"media/ikea-mujoco/stegesund.jpg?v=928bebaf","video":"media/ikea-mujoco/stegesund.mp4?v=1b868ce2","motion":true,"aspect":1.778,"seconds":32},{"task":"svallet","title":"Svallet","poster":"media/ikea-mujoco/svallet.jpg?v=ad182e72","video":"media/ikea-mujoco/svallet.mp4?v=846d2a2c","motion":true,"aspect":1.778,"seconds":24},{"task":"lack","title":"Lack","poster":"media/ikea-mujoco/lack.jpg?v=2d957909","video":"media/ikea-mujoco/lack.mp4?v=1d42c53c","motion":true,"aspect":1.778,"seconds":22}],"teaser":{"poster":"media/roboarc/_teaser.jpg?v=3893b0e7","video":"media/roboarc/_teaser.mp4?v=df73c006","cover":true,"aspect":1.778,"seconds":46},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"robodojo","name":"RoboDojo","code":"RDJ","origin":"adapted","example":false,"blurb":"Bimanual manipulation over five capabilities: generalization, memory, precision, long horizons, open tasks.","description":"The simulation part of RoboDojo's benchmark for generalist manipulation policies: two ARX X5 arms with parallel grippers at a table. All 42 tasks over five capability dimensions (generalization, memory, precision, long horizon and open tasks) plus the 12 randomised variants of the generalization tasks, with other models, lighting, backgrounds and clutter. Each task is one frozen evaluation layout; the six whose physics does not replay bit for bit are standard only.","simulator":"Isaac Lab 2.3 / Isaac Sim 5.1","links":{"repository":"https://github.com/RoboDojo-Benchmark/RoboDojo","paper":"https://arxiv.org/abs/2607.04434","project_page":"https://robodojo-benchmark.com/"},"robots":[{"id":"arx-x5-bimanual","name":"Two ARX X5 arms","body":"bimanual"}],"body":"bimanual","bodies":{"bimanual":54},"instances":54,"task_dirs":102,"modes":{"privileged":48,"standard":54},"oracle":{"full":0,"partial":0,"none":54},"protocols":["eai-standard/2.2"],"budgets_min":[60],"on_site":false,"preview":[{"task":"stack_blocks_by_language","title":"Stack blocks by language","poster":"media/robodojo/stack_blocks_by_language.jpg?v=942f8ee7","video":"media/robodojo/stack_blocks_by_language.mp4?v=79018812","motion":true,"aspect":1.778,"seconds":46},{"task":"match_and_pick_from_conveyor","title":"Match and pick from conveyor","poster":"media/robodojo/match_and_pick_from_conveyor.jpg?v=51b48f49","video":"media/robodojo/match_and_pick_from_conveyor.mp4?v=810744f1","motion":true,"aspect":1.778,"seconds":14},{"task":"deposit_coin","title":"Deposit coin","poster":"media/robodojo/deposit_coin.jpg?v=1cc7029a","video":"media/robodojo/deposit_coin.mp4?v=0f0e09a5","motion":true,"aspect":1.778,"seconds":21},{"task":"general_pickup","title":"General pickup","poster":"media/robodojo/general_pickup.jpg?v=71bca258","video":"media/robodojo/general_pickup.mp4?v=034122e4","motion":true,"aspect":1.778,"seconds":48},{"task":"align_blocks","title":"Align blocks","poster":"media/robodojo/align_blocks.jpg?v=aaa052c6","video":"media/robodojo/align_blocks.mp4?v=19607913","motion":false,"aspect":1.778,"seconds":7},{"task":"arrange_largest_number","title":"Arrange largest number","poster":"media/robodojo/arrange_largest_number.jpg?v=508e6f23","video":"media/robodojo/arrange_largest_number.mp4?v=3c85be68","motion":false,"aspect":1.778,"seconds":7},{"task":"arrange_largest_number_random","title":"Arrange largest number random","poster":"media/robodojo/arrange_largest_number_random.jpg?v=6eff8f76","video":"media/robodojo/arrange_largest_number_random.mp4?v=8a1d616c","motion":false,"aspect":1.778,"seconds":7},{"task":"build_tower","title":"Build tower","poster":"media/robodojo/build_tower.jpg?v=adb319e5","video":"media/robodojo/build_tower.mp4?v=bcad4b11","motion":false,"aspect":1.778,"seconds":7}],"teaser":{"poster":"media/robodojo/_teaser.jpg?v=942f8ee7","video":"media/robodojo/_teaser.mp4?v=de6e5a5e","cover":true,"aspect":1.778,"seconds":23},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"pandadyn","name":"Panda Dynamic Scenes","code":"PDS","origin":"ours","example":false,"blurb":"Catch, strike, carry and hand over objects that move by themselves.","description":"25 continuous, multi-stage dynamic tasks for one or two Franka Panda arms with task tools (a plate, a cup, a paddle, a scoop): in every task something moves by itself, such as a tossed ball, a sliding puck, a conveyor, a pendulum or a moving cart, and the robot must catch, strike, carry, hand over, flip or release at the right moment, several times in one episode.","simulator":"ManiSkill 3 / SAPIEN (CPU PhysX at 1000 Hz)","robots":[{"id":"franka-panda","name":"Franka Panda","body":"arm"},{"id":"franka-panda-bimanual","name":"Two Franka Panda arms","body":"bimanual"}],"body":"bimanual","bodies":{"bimanual":13,"arm":12},"instances":25,"task_dirs":69,"modes":{"privileged":25,"standard":25,"standard-realtime":19},"oracle":{"full":25,"partial":0,"none":0},"protocols":["eai-standard/2.1","eai-standard/2.3"],"budgets_min":[60],"on_site":false,"preview":[{"task":"d01_reciprocal_cup_relay","title":"D01 reciprocal cup relay","poster":"media/pandadyn/d01_reciprocal_cup_relay.jpg?v=3d2e3031","video":"media/pandadyn/d01_reciprocal_cup_relay.mp4?v=f0453419","motion":true,"aspect":1.778,"seconds":16},{"task":"d09_lever_launch_catch_reload","title":"D09 lever launch catch reload","poster":"media/pandadyn/d09_lever_launch_catch_reload.jpg?v=15c1f96a","video":"media/pandadyn/d09_lever_launch_catch_reload.mp4?v=87956277","motion":true,"aspect":1.778,"seconds":17},{"task":"d02_strike_catch_merge","title":"D02 strike catch merge","poster":"media/pandadyn/d02_strike_catch_merge.jpg?v=d936c498","video":"media/pandadyn/d02_strike_catch_merge.mp4?v=3cdd24fa","motion":true,"aspect":1.778,"seconds":19},{"task":"d05_moving_ramp_relay","title":"D05 moving ramp relay","poster":"media/pandadyn/d05_moving_ramp_relay.jpg?v=da251c14","video":"media/pandadyn/d05_moving_ramp_relay.mp4?v=e0aa19a3","motion":true,"aspect":1.778,"seconds":25},{"task":"d10_moving_grasp_handover","title":"D10 moving grasp handover","poster":"media/pandadyn/d10_moving_grasp_handover.jpg?v=57d7883b","video":"media/pandadyn/d10_moving_grasp_handover.mp4?v=2aeccee4","motion":true,"aspect":1.778,"seconds":14},{"task":"s09_toss_recatch_delivery","title":"S09 toss recatch delivery","poster":"media/pandadyn/s09_toss_recatch_delivery.jpg?v=3a6d9b52","video":"media/pandadyn/s09_toss_recatch_delivery.mp4?v=19ff64cc","motion":true,"aspect":1.778,"seconds":13},{"task":"s12_moving_bridge_puck_relay","title":"S12 moving bridge puck relay","poster":"media/pandadyn/s12_moving_bridge_puck_relay.jpg?v=17775754","video":"media/pandadyn/s12_moving_bridge_puck_relay.mp4?v=fcd75b1d","motion":true,"aspect":1.778,"seconds":32},{"task":"d03_cooperative_puck_course","title":"D03 cooperative puck course","poster":"media/pandadyn/d03_cooperative_puck_course.jpg?v=ca8c1d45","video":"media/pandadyn/d03_cooperative_puck_course.mp4?v=329b2685","motion":true,"aspect":1.778,"seconds":24}],"teaser":{"poster":"media/pandadyn/_teaser.jpg?v=3d2e3031","video":"media/pandadyn/_teaser.mp4?v=f0453419","cover":true,"aspect":1.778,"seconds":16},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"bigym2","name":"BiGym 2.0","code":"BG2","origin":"adapted","example":false,"blurb":"Household loco-manipulation on a walking Unitree G1.","description":"Household tasks such as a dishwasher, cupboards and drawers for a Unitree G1 with two grippers, kept balanced and walking by NVIDIA's GR00T-WBC policy. The agent commands what a real G1 running that controller takes: walking velocity, pelvis height, torso lean, arm joints and grippers. 13 families, each also in standard-realtime mode, where the world keeps running while the agent thinks.","simulator":"MuJoCo 3.8 + NVIDIA GR00T-WBC","links":{"repository":"https://github.com/swirl-uk/BiGym2","paper":"https://arxiv.org/abs/2610.07594","project_page":"https://bigym2.github.io/"},"robots":[{"id":"unitree-g1-dex1","name":"Unitree G1 with grippers","body":"humanoid"}],"body":"humanoid","bodies":{"humanoid":13},"instances":13,"task_dirs":39,"modes":{"privileged":13,"standard":13,"standard-realtime":13},"oracle":{"full":0,"partial":0,"none":13},"protocols":["eai-standard/2.1","eai-standard/2.3"],"budgets_min":[60],"on_site":false,"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"yamlab","name":"YAMLab","code":"YAM","origin":"adapted","example":false,"blurb":"Two I2RT YAM arms: set a pot on a cooktop, hand a mug over and hang it.","description":"Bimanual manipulation with two I2RT YAM arms at a workstation table, YAMLab's two tasks: lift a pot with both grippers and set it upright on a cooktop, and lift a mug with one gripper, hand it to the other in the air and hang it on a mug tree. Five object pairs each, placed as in YAMLab's human teleoperation demonstrations.","simulator":"Isaac Lab / Isaac Sim 5.1","links":{"repository":"https://github.com/ARISE-Initiative/yamlab"},"robots":[{"id":"i2rt-yam-bimanual","name":"Two I2RT YAM arms","body":"bimanual"}],"body":"bimanual","bodies":{"bimanual":10},"instances":10,"task_dirs":20,"modes":{"privileged":10,"standard":10},"oracle":{"full":10,"partial":0,"none":0},"protocols":["eai-standard/2.2"],"budgets_min":[60],"on_site":false,"preview":[{"task":"hang_mug_on_tree_i15","title":"Hang mug on tree i15","poster":"media/yamlab/hang_mug_on_tree_i15.jpg?v=5eb82628","video":"media/yamlab/hang_mug_on_tree_i15.mp4?v=357b90da","motion":true,"aspect":1.778,"seconds":23},{"task":"put_pot_on_cooktop_i07","title":"Put pot on cooktop i07","poster":"media/yamlab/put_pot_on_cooktop_i07.jpg?v=68d52fcf","video":"media/yamlab/put_pot_on_cooktop_i07.mp4?v=e9536345","motion":true,"aspect":1.778,"seconds":14},{"task":"hang_mug_on_tree_i13","title":"Hang mug on tree i13","poster":"media/yamlab/hang_mug_on_tree_i13.jpg?v=784210e1","video":"media/yamlab/hang_mug_on_tree_i13.mp4?v=807ff3f8","motion":true,"aspect":1.778,"seconds":20},{"task":"hang_mug_on_tree_i21","title":"Hang mug on tree i21","poster":"media/yamlab/hang_mug_on_tree_i21.jpg?v=c81f67d9","video":"media/yamlab/hang_mug_on_tree_i21.mp4?v=e0819519","motion":true,"aspect":1.778,"seconds":22},{"task":"put_pot_on_cooktop_i08","title":"Put pot on cooktop i08","poster":"media/yamlab/put_pot_on_cooktop_i08.jpg?v=18a333d4","video":"media/yamlab/put_pot_on_cooktop_i08.mp4?v=9eb261a0","motion":true,"aspect":1.778,"seconds":13},{"task":"hang_mug_on_tree_i10","title":"Hang mug on tree i10","poster":"media/yamlab/hang_mug_on_tree_i10.jpg?v=5d1d9093","video":"media/yamlab/hang_mug_on_tree_i10.mp4?v=407ba130","motion":true,"aspect":1.778,"seconds":21},{"task":"hang_mug_on_tree_i20","title":"Hang mug on tree i20","poster":"media/yamlab/hang_mug_on_tree_i20.jpg?v=5f145dbf","video":"media/yamlab/hang_mug_on_tree_i20.mp4?v=da42e40d","motion":true,"aspect":1.778,"seconds":21},{"task":"put_pot_on_cooktop_i03","title":"Put pot on cooktop i03","poster":"media/yamlab/put_pot_on_cooktop_i03.jpg?v=4b0c19e0","video":"media/yamlab/put_pot_on_cooktop_i03.mp4?v=182f0621","motion":true,"aspect":1.778,"seconds":15}],"teaser":{"poster":"media/yamlab/_teaser.jpg?v=5eb82628","video":"media/yamlab/_teaser.mp4?v=357b90da","cover":true,"aspect":1.778,"seconds":23},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"libero-max","name":"LIBERO-MAX","code":"LBX","origin":"adapted","example":false,"blurb":"LIBERO tasks with one change in the middle of the task.","description":"64 LIBERO-MAX cases for a Franka Panda, built on LIBERO-PRO variants of LIBERO's kitchen and living-room tasks: pick-and-place, stove knobs and a drawer to close, each with one event in the middle of the task, such as an object that moves, a distractor that drops in, an obstacle that appears or a visual theme that switches. The robot has to notice the change and still finish the goal. Each case is one frozen scene, in both modes.","simulator":"robosuite 1.4 / MuJoCo 3.2.6","links":{"repository":"https://github.com/liberomax/LIBERO-MAX","paper":"https://arxiv.org/abs/2609.36518","project_page":"https://liberomax.github.io/"},"robots":[{"id":"franka-panda","name":"Franka Panda","body":"arm"}],"body":"arm","bodies":{"arm":64},"instances":64,"task_dirs":128,"modes":{"privileged":64,"standard":64},"oracle":{"full":64,"partial":0,"none":0},"protocols":["eai-standard/2.1"],"budgets_min":[30],"on_site":false,"preview":[{"task":"goal_akita_black_bowl_in_wooden_cabinet_top_target_relocation","title":"Goal akita black bowl in wooden cabinet top target relocation","poster":"media/libero-max/goal_akita_black_bowl_in_wooden_cabinet_top_target_relocation.jpg?v=93abe185","video":"media/libero-max/goal_akita_black_bowl_in_wooden_cabinet_top_target_relocation.mp4?v=48aa9db2","motion":true,"aspect":1.778,"seconds":14},{"task":"goal_akita_black_bowl_on_flat_stove_cook_target_relocation","title":"Goal akita black bowl on flat stove cook target relocation","poster":"media/libero-max/goal_akita_black_bowl_on_flat_stove_cook_target_relocation.jpg?v=a46a399b","video":"media/libero-max/goal_akita_black_bowl_on_flat_stove_cook_target_relocation.mp4?v=20720695","motion":true,"aspect":1.778,"seconds":8},{"task":"goal_akita_black_bowl_on_plate_receptacle_relocation","title":"Goal akita black bowl on plate receptacle relocation","poster":"media/libero-max/goal_akita_black_bowl_on_plate_receptacle_relocation.jpg?v=7fb99aa0","video":"media/libero-max/goal_akita_black_bowl_on_plate_receptacle_relocation.mp4?v=3ba6418d","motion":true,"aspect":1.778,"seconds":8},{"task":"goal_akita_black_bowl_on_wooden_cabinet_top_obstacle_insertion","title":"Goal akita black bowl on wooden cabinet top obstacle insertion","poster":"media/libero-max/goal_akita_black_bowl_on_wooden_cabinet_top_obstacle_insertion.jpg?v=05eb7d0d","video":"media/libero-max/goal_akita_black_bowl_on_wooden_cabinet_top_obstacle_insertion.mp4?v=66fa94f5","motion":true,"aspect":1.778,"seconds":10},{"task":"goal_cream_cheese_on_akita_black_bowl_obstacle_insertion","title":"Goal cream cheese on akita black bowl obstacle insertion","poster":"media/libero-max/goal_cream_cheese_on_akita_black_bowl_obstacle_insertion.jpg?v=97cc834d","video":"media/libero-max/goal_cream_cheese_on_akita_black_bowl_obstacle_insertion.mp4?v=bd0765e2","motion":true,"aspect":1.778,"seconds":15},{"task":"goal_flat_stove_off_visual_theme_switch","title":"Goal flat stove off visual theme switch","poster":"media/libero-max/goal_flat_stove_off_visual_theme_switch.jpg?v=6e4e6e7b","video":"media/libero-max/goal_flat_stove_off_visual_theme_switch.mp4?v=20ae300b","motion":true,"aspect":1.778,"seconds":6},{"task":"goal_flat_stove_on_illumination_switch","title":"Goal flat stove on illumination switch","poster":"media/libero-max/goal_flat_stove_on_illumination_switch.jpg?v=228105cd","video":"media/libero-max/goal_flat_stove_on_illumination_switch.mp4?v=8b4fa2ac","motion":true,"aspect":1.778,"seconds":6},{"task":"goal_flat_stove_on_illumination_switch_i01","title":"Goal flat stove on illumination switch i01","poster":"media/libero-max/goal_flat_stove_on_illumination_switch_i01.jpg?v=f36ab77e","video":"media/libero-max/goal_flat_stove_on_illumination_switch_i01.mp4?v=7e0ee9ff","motion":true,"aspect":1.778,"seconds":6}],"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"roboolympics","name":"RoboOlympics","code":"OLY","origin":"ours","example":false,"blurb":"Fifty humanoid sports at official sizes, judged by a referee, and real-time matches.","description":"Fifty single-robot sports for a Unitree G1 or the adult-size H1-2, in their official settings: sprint, hurdles, jumps, a penalty against a keeper, a serve over a 2.43 m net, a shot at a 3.05 m ring, a table-tennis return, a 7.26 kg shot put, bowling, a giant swing on the high bar. A referee judges from the simulator state whether the sport was done. Nothing is scaled down for the robot, most tasks have no known solution, and the agent may use the robot's walking policy, drive every joint itself, or download public controllers. Eight more tasks come with their own sport service: matches against a scripted team in real time (table tennis, badminton, humanoid soccer with Booster T1 robots, wheeled small-size soccer) and two skill-building tasks.","simulator":"MuJoCo 3.3.7 (CPU); the matches on their own sport services","links":{"project_page":"https://olympics.embodied-first-exam.ai"},"robots":[{"id":"booster-t1","name":"Booster T1","body":"humanoid"},{"id":"foosball-rods","name":"Foosball rods","body":"arm"},{"id":"franka-panda","name":"Franka Panda","body":"arm"},{"id":"ssl-robot","name":"RoboCup Small Size robots","body":"mobile"},{"id":"unitree-g1","name":"Unitree G1","body":"humanoid"},{"id":"unitree-h1-2","name":"Unitree H1-2","body":"humanoid"}],"body":"humanoid","bodies":{"humanoid":55,"arm":12,"mobile":2},"instances":69,"task_dirs":138,"modes":{"privileged":56,"standard":52,"standard-realtime":30},"oracle":{"full":14,"partial":17,"none":38},"protocols":["eai-standard-realtime/1.0","eai-standard-turn/1.0","eai-standard/2.1","eai-standard/2.3"],"budgets_min":[120,240],"on_site":false,"preview":[{"task":"dunk","title":"Dunk","poster":"media/roboolympics/dunk.jpg?v=03e8838d","video":"media/roboolympics/dunk.mp4?v=2c0401b6","motion":true,"aspect":1.778,"seconds":24}],"teaser":{"poster":"media/roboolympics/_teaser.jpg?v=03e8838d","video":"media/roboolympics/_teaser.mp4?v=2c0401b6","cover":true,"aspect":1.778,"seconds":24},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"ambench","name":"AM-Bench","code":"AMB","origin":"adapted","example":false,"blurb":"Aerial manipulation: a flying hexarotor with an arm presses, pulls, turns and picks.","description":"Nine of AM-Bench's aerial manipulation tasks for its FA-Hexa, a fully actuated hexarotor that can hold any attitude while hovering, carrying a four-joint arm with a two-finger gripper: press a button and push a slider on a tilted wall, turn a valve, open a door, insert a peg, hold an inspection point high on a warehouse wall, harvest a lemon and lift a can onto a cabinet shelf. One frozen instance each; the lever task is standard only.","simulator":"Isaac Lab / Isaac Sim 5.1","links":{"repository":"https://github.com/ambench/ambench","paper":"https://arxiv.org/abs/2609.00641","project_page":"https://ambench.github.io/"},"robots":[{"id":"ambench-fahexa","name":"AM-Bench FA-Hexa (hexarotor with an arm)","body":"aerial"}],"body":"aerial","bodies":{"aerial":9},"instances":9,"task_dirs":17,"modes":{"privileged":8,"standard":9},"oracle":{"full":9,"partial":0,"none":0},"protocols":["eai-standard/2.2"],"budgets_min":[60],"on_site":false,"preview":[{"task":"cabinet_pick_place","title":"Cabinet pick place","poster":"media/ambench/cabinet_pick_place.jpg?v=76431943","video":"media/ambench/cabinet_pick_place.mp4?v=958b7603","motion":true,"aspect":1.778,"seconds":24},{"task":"lemon_harvesting","title":"Lemon harvesting","poster":"media/ambench/lemon_harvesting.jpg?v=6b2281e1","video":"media/ambench/lemon_harvesting.mp4?v=026a00da","motion":true,"aspect":1.778,"seconds":16},{"task":"peg_in_hole","title":"Peg in hole","poster":"media/ambench/peg_in_hole.jpg?v=0dbe534b","video":"media/ambench/peg_in_hole.mp4?v=5105e462","motion":true,"aspect":1.778,"seconds":10},{"task":"rotate_valve","title":"Rotate valve","poster":"media/ambench/rotate_valve.jpg?v=9f003742","video":"media/ambench/rotate_valve.mp4?v=77c8da56","motion":true,"aspect":1.778,"seconds":9},{"task":"pull_lever","title":"Pull lever","poster":"media/ambench/pull_lever.jpg?v=60c74b47","video":"media/ambench/pull_lever.mp4?v=9d3299a9","motion":true,"aspect":1.778,"seconds":9},{"task":"open_door","title":"Open door","poster":"media/ambench/open_door.jpg?v=9ac5dc81","video":"media/ambench/open_door.mp4?v=14043aa2","motion":true,"aspect":1.778,"seconds":5},{"task":"ndt","title":"Ndt","poster":"media/ambench/ndt.jpg?v=33d0d1f0","video":"media/ambench/ndt.mp4?v=b107f8ff","motion":true,"aspect":1.778,"seconds":11},{"task":"press_button","title":"Press button","poster":"media/ambench/press_button.jpg?v=027e2049","video":"media/ambench/press_button.mp4?v=b2f57d7b","motion":true,"aspect":1.778,"seconds":10}],"teaser":{"poster":"media/ambench/_teaser.jpg?v=76431943","video":"media/ambench/_teaser.mp4?v=958b7603","cover":true,"aspect":1.778,"seconds":24},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"robosafe","name":"RoboSafeBench","code":"RSB","origin":"adapted","example":false,"blurb":"Do what is harmless, refuse what would hurt someone, and say why.","description":"Safety for an agent that operates a robot for a person. Every family pairs requests whose literal execution ends in harm (an injury, a fire, poison, a shock, something broken) with harmless twins of the same kind. The verdict comes from what physically happened in the simulation, read by written hazard rules, and from how the episode ended, never from the agent's text. A UR5e, a Franka FR3, two YAM arms, a TidyBot++ and a walking G1, with low-level control; standard mode, plus a real-time set.","simulator":"MuJoCo 3.8.1 (CPU)","robots":[{"id":"franka-robotiq-2f85","name":"Franka + Robotiq 2F-85","body":"arm"},{"id":"i2rt-yam-bimanual","name":"Two I2RT YAM arms","body":"bimanual"},{"id":"tidybot-plus-plus","name":"TidyBot++","body":"mobile"},{"id":"unitree-g1-dex1","name":"Unitree G1 with grippers","body":"humanoid"},{"id":"ur5e-robotiq-2f85","name":"UR5e + Robotiq 2F-85","body":"arm"}],"body":"arm","bodies":{"arm":404,"bimanual":53,"mobile":56,"humanoid":24},"instances":537,"task_dirs":566,"modes":{"privileged":0,"standard":537,"standard-realtime":29},"oracle":{"full":537,"partial":0,"none":0},"protocols":["eai-standard/2.1","eai-standard/2.3"],"budgets_min":[60],"on_site":false,"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"dexhand","name":"DexHand","code":"DXH","origin":"ours","example":false,"blurb":"Five-finger hands: reorient, regrasp, fold paper, sweep and build a hamburger.","description":"Ten dexterous-hand tasks, one frozen scene each: six with a Sharpa Wave hand on a wrist stage (cube layers, a held joystick, a pen regrasp, two objects, a rotary gait, a cube reorientation), assembling a hamburger, folding paper in half, and two with a hand on each of two Tianji Marvin arms (sweeping into a dustpan, moving a pot to a shelf). No assisted grasp: objects move only through contact and gravity. Every task also runs with a Wuji hand.","simulator":"MuJoCo 3.4.0 (CPU)","robots":[{"id":"marvin-bimanual-sharpa-wave","name":"Two Tianji Marvin arms with Sharpa Wave hands","body":"hand"},{"id":"sharpa-wave-floating","name":"Sharpa Wave hand on a wrist stage","body":"hand"},{"id":"sharpa-wave-floating-bimanual","name":"Two Sharpa Wave hands on wrist stages","body":"hand"}],"body":"hand","bodies":{"hand":30},"instances":30,"task_dirs":60,"modes":{"privileged":30,"standard":30},"oracle":{"full":30,"partial":0,"none":0},"protocols":["eai-standard/2.2"],"budgets_min":[60],"on_site":false,"preview":[{"task":"assemble_hamburger","title":"Assemble hamburger","poster":"media/dexhand/assemble_hamburger.jpg?v=4feee158","video":"media/dexhand/assemble_hamburger.mp4?v=04394b97","motion":true,"aspect":1.778,"seconds":22},{"task":"sweep_dustpan","title":"Sweep dustpan","poster":"media/dexhand/sweep_dustpan.jpg?v=f70f35e8","video":"media/dexhand/sweep_dustpan.mp4?v=d0022203","motion":true,"aspect":1.778,"seconds":25},{"task":"pot_shelf_transfer","title":"Pot shelf transfer","poster":"media/dexhand/pot_shelf_transfer.jpg?v=2ffb10c1","video":"media/dexhand/pot_shelf_transfer.mp4?v=72597d66","motion":true,"aspect":1.778,"seconds":20},{"task":"paper_half_fold","title":"Paper half fold","poster":"media/dexhand/paper_half_fold.jpg?v=9936822a","video":"media/dexhand/paper_half_fold.mp4?v=c605f902","motion":true,"aspect":1.778,"seconds":17},{"task":"essential_rotary_gait","title":"Essential rotary gait","poster":"media/dexhand/essential_rotary_gait.jpg?v=84802603","video":"media/dexhand/essential_rotary_gait.mp4?v=63e84a17","motion":true,"aspect":1.778,"seconds":29},{"task":"batter_whisk","title":"Batter whisk","poster":"media/dexhand/batter_whisk.jpg?v=09757923","video":"media/dexhand/batter_whisk.mp4?v=a073649a","motion":true,"aspect":1.778,"seconds":21},{"task":"binder_page_load","title":"Binder page load","poster":"media/dexhand/binder_page_load.jpg?v=9f19a3c5","video":"media/dexhand/binder_page_load.mp4?v=4087618b","motion":true,"aspect":1.778,"seconds":22},{"task":"book_page_turn","title":"Book page turn","poster":"media/dexhand/book_page_turn.jpg?v=90ec4ed3","video":"media/dexhand/book_page_turn.mp4?v=221e8900","motion":true,"aspect":1.778,"seconds":12}],"teaser":{"poster":"media/dexhand/_teaser.jpg?v=4feee158","video":"media/dexhand/_teaser.mp4?v=04394b97","cover":true,"aspect":1.778,"seconds":22},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"ispatialgym","name":"iSpatialGym","code":"ISG","origin":"adapted","example":false,"blurb":"The goal is a picture: make the scene look like it.","description":"Spatial intelligence on a mobile two-armed robot in house-scale scenes. Each task gives only a goal image, the head camera's view once the task is done, and the robot has to make the scene match it: an object's pose, a relation between two objects, an articulation, or its own viewpoint. No sentence names the objects.","simulator":"OmniGibson (BEHAVIOR-1K fork) on Isaac Sim 5.1","links":{"repository":"https://github.com/kmy17518/iSpatialGym"},"robots":[{"id":"r1pro","name":"Galaxea R1 Pro","body":"mobile"}],"body":"mobile","bodies":{"mobile":12},"instances":12,"task_dirs":24,"modes":{"privileged":12,"standard":12},"oracle":{"full":0,"partial":0,"none":12},"protocols":["eai-standard/2.2"],"budgets_min":[60],"on_site":false,"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}},{"id":"robowits","name":"RoboWits","code":"RWT","origin":"adapted","example":false,"blurb":"Tabletop tasks that need a physical insight, like a ruler to align things or a funnel to pour.","description":"30 seed tabletop tasks for a fixed bimanual robot, each requiring a physical insight rather than a longer motion: use a ruler to align, a dustpan to sweep, a funnel to pour, an inclined plane to move a box. 163 published mutations vary the scene while keeping the parent task's success predicate. Scored by the task's own success check.","edition":"code release 2026-06-04 (commit 9cc30ae)","simulator":"Genesis 0.4.7 (gs_gym)","links":{"paper":"https://arxiv.org/abs/2605.30326","project_page":"https://umass-embodied-agi.github.io/RoboWits/","repository":"https://github.com/UMass-Embodied-AGI/RoboWits","dataset":"https://huggingface.co/datasets/XHRlyb2001/RoboWits_lerobot_dataset","assets":"https://huggingface.co/datasets/XHRlyb2001/RoboWits_assets"},"robots":[{"id":"marvin-bimanual","name":"Marvin","body":"bimanual"}],"body":"bimanual","bodies":{"bimanual":15},"instances":15,"task_dirs":30,"modes":{"privileged":15,"standard":15},"oracle":{"full":0,"partial":2,"none":13},"protocols":["eai-standard/2.2"],"budgets_min":[60],"tags":{"planning-reasoning":15,"control-coordination":15,"manipulation":15,"perception-understanding":2},"on_site":true,"preview":[{"task":"ball_onto_tower","title":"Ball Onto Tower","cover":true,"poster":"media/robowits/ball_onto_tower.jpg?v=2e534369","video":"media/robowits/ball_onto_tower.mp4?v=66ac5101","motion":true,"aspect":1.778,"seconds":37},{"task":"stand_pages","title":"Stand Pages","poster":"media/robowits/stand_pages.jpg?v=4ba023b1","video":"media/robowits/stand_pages.mp4?v=714eb045","motion":true,"aspect":1.778,"seconds":20},{"task":"ball_into_bottle","title":"Ball Into Bottle","poster":"media/robowits/ball_into_bottle.jpg?v=2237b953","video":"media/robowits/ball_into_bottle.mp4?v=b0097f3c","motion":true,"aspect":1.778,"seconds":17},{"task":"hold_cup","title":"Hold Cup","poster":"media/robowits/hold_cup.jpg?v=8a1f5220","video":"media/robowits/hold_cup.mp4?v=524e0ec3","motion":true,"aspect":1.778,"seconds":38},{"task":"raise_platform","title":"Raise Platform","poster":"media/robowits/raise_platform.jpg?v=8fa51128","video":"media/robowits/raise_platform.mp4?v=16fc1e2f","motion":true,"aspect":1.778,"seconds":48},{"task":"dominos","title":"Dominos","poster":"media/robowits/dominos.jpg?v=55ae1e4c","video":"media/robowits/dominos.mp4?v=76c71b5f","motion":true,"aspect":1.778,"seconds":13},{"task":"move_cube","title":"Move Cube","poster":"media/robowits/move_cube.jpg?v=bd3425b2","video":"media/robowits/move_cube.mp4?v=3e7d4cbb","motion":true,"aspect":1.778,"seconds":19},{"task":"align_blocks","title":"Align Blocks","poster":"media/robowits/align_blocks.jpg?v=d9652a8a","video":"media/robowits/align_blocks.mp4?v=17aec89a","motion":true,"aspect":1.778,"seconds":14}],"results":{"codex-gpt6_luna-xhigh":{"privileged":{"done":15,"success":11,"trials":15},"standard":{"done":15,"success":8,"trials":15}},"codex-gpt6_astra-high":{"privileged":{"done":14,"success":13,"trials":14},"standard":{"done":13,"success":10,"trials":14}},"codex-gpt6_sol-high":{"privileged":{"done":10,"success":10,"trials":11},"standard":{"done":9,"success":5,"trials":11}},"codex-gpt6_1_sol-high":{"privileged":{"done":10,"success":10,"trials":11},"standard":{"done":9,"success":7,"trials":11}}},"solved":{"privileged":14,"standard":12},"attempted":{"privileged":15,"standard":15},"cost":{"codex-gpt6_luna-xhigh":{"est_usd":6.26,"est_n":30,"agent_s":88916,"agent_n":30},"codex-gpt6_astra-high":{"est_usd":150.32,"est_n":27,"agent_s":23078,"agent_n":27},"codex-gpt6_sol-high":{"est_usd":54.56,"est_n":19,"agent_s":34977,"agent_n":19},"codex-gpt6_1_sol-high":{"est_usd":27.87,"est_n":19,"agent_s":26114,"agent_n":19}}},{"id":"maniguard","name":"ManiGuard","code":"MGD","origin":"adapted","example":false,"blurb":"Manipulation safety. A task counts only if no safety rule was broken on the way.","description":"A manipulation safety benchmark from the NU IDEAS Lab: tabletop tasks for a Franka Panda with long fin-ray fingers, each with a safety specification in linear temporal logic over physics predicates (dropped, upright, on top of, closed, touching), checked after every control step. A task counts only if the goal holds and no rule was ever broken.","simulator":"BEHAVIOR-1K v3.7.2 (OmniGibson 3.7.2 on Isaac Sim 4.5)","links":{"paper":"https://arxiv.org/abs/2608.17386","repository":"https://github.com/NU-IDEAS-Lab/ManiGuard","project_page":"https://nu-ideas-lab.github.io/ManiGuard/"},"robots":[{"id":"franka-panda-longfinger","name":"Franka Panda","body":"arm"}],"body":"arm","bodies":{"arm":30},"instances":30,"task_dirs":60,"modes":{"privileged":30,"standard":30},"oracle":{"full":14,"partial":0,"none":16},"protocols":["eai-standard/2.2"],"budgets_min":[60],"on_site":false,"preview":[{"task":"cabinet_pickup_i22","title":"Cabinet pickup i22","cover":true,"poster":"media/maniguard/cabinet_pickup_i22.jpg?v=48ae2f29","video":"media/maniguard/cabinet_pickup_i22.mp4?v=4f6f2999","motion":true,"aspect":1.778,"seconds":41},{"task":"cabinet_pickup_i19","title":"Cabinet pickup i19","poster":"media/maniguard/cabinet_pickup_i19.jpg?v=f538332d","video":"media/maniguard/cabinet_pickup_i19.mp4?v=dd480d91","motion":true,"aspect":1.778,"seconds":39},{"task":"dusty_transfer_i24","title":"Dusty transfer i24","poster":"media/maniguard/dusty_transfer_i24.jpg?v=202444cd","video":"media/maniguard/dusty_transfer_i24.mp4?v=d6aef578","motion":true,"aspect":1.778,"seconds":18},{"task":"lid_transport_i07","title":"Lid transport i07","poster":"media/maniguard/lid_transport_i07.jpg?v=0397e229","video":"media/maniguard/lid_transport_i07.mp4?v=4d54d541","motion":true,"aspect":1.778,"seconds":33},{"task":"stack_retrieve_i08","title":"Stack retrieve i08","poster":"media/maniguard/stack_retrieve_i08.jpg?v=b61c26d9","video":"media/maniguard/stack_retrieve_i08.mp4?v=79de84da","motion":true,"aspect":1.778,"seconds":28},{"task":"cabinet_pickup_i20","title":"Cabinet pickup i20","poster":"media/maniguard/cabinet_pickup_i20.jpg?v=7e767808","video":"media/maniguard/cabinet_pickup_i20.mp4?v=706ff5ad","motion":true,"aspect":1.778,"seconds":37},{"task":"cabinet_pickup_i26","title":"Cabinet pickup i26","poster":"media/maniguard/cabinet_pickup_i26.jpg?v=58e2a4e6","video":"media/maniguard/cabinet_pickup_i26.mp4?v=d1f25c9a","motion":true,"aspect":1.778,"seconds":25},{"task":"cabinet_pickup_i29","title":"Cabinet pickup i29","poster":"media/maniguard/cabinet_pickup_i29.jpg?v=62bf3326","video":"media/maniguard/cabinet_pickup_i29.mp4?v=7a69f8af","motion":true,"aspect":1.778,"seconds":28}],"teaser":{"poster":"media/maniguard/_teaser.jpg?v=f538332d","video":"media/maniguard/_teaser.mp4?v=dd480d91","cover":true,"aspect":1.778,"seconds":39},"solved":{"privileged":0,"standard":0},"attempted":{"privileged":0,"standard":0}}],"bodies":{"arm":"One arm","bimanual":"Two arms","mobile":"Mobile manipulator","humanoid":"Humanoid","quadruped":"Quadruped","hand":"Dexterous hand","musculoskeletal":"Musculoskeletal","aerial":"Aerial manipulator"},"taxonomy":[{"id":"capability","name":"Capability","description":"What the agent has to be able to do to solve the task. Take every one the goal requires.","tags":[{"id":"perception-understanding","name":"Perception & Understanding","description":"Perception, object/state recognition, semantic grounding, spatial understanding."},{"id":"planning-reasoning","name":"Planning & Reasoning","description":"Task planning, sequencing, tool/affordance reasoning, geometric reasoning."},{"id":"control-coordination","name":"Control & Coordination","description":"Manipulation/control, contact-rich execution, locomotion control, multi-arm or whole-body coordination."},{"id":"feedback-adaptation","name":"Feedback & Adaptation","description":"Closed-loop correction, failure recovery, online adaptation, exploration based on feedback."}]},{"id":"domain","name":"Task Domain","description":"What kind of task it is: where the robot works and what its body has to do. Take every one that applies.","tags":[{"id":"manipulation","name":"Manipulation","description":"Fixed-base object manipulation, grasping, pushing, insertion, tool use, articulated/deformable object interaction."},{"id":"locomotion-stability","name":"Locomotion & Stability","description":"Walking, running, balancing, terrain traversal, dynamic stabilization."},{"id":"navigation-exploration","name":"Navigation & Exploration","description":"Navigation, target search, exploration, spatial memory."},{"id":"mobile-whole-body-manipulation","name":"Mobile / Whole-body Manipulation","description":"Tasks combining navigation/locomotion with manipulation or requiring coordinated whole-body motion."},{"id":"interaction-collaboration","name":"Interaction & Collaboration","description":"Human-robot interaction, multi-robot collaboration, handover, cooperative tasks."}]}],"runs":[{"id":"codex-gpt6_luna-xhigh","model":"GPT-6 Luna","harness":"Codex","effort":"xhigh","route":"ChatGPT login","label":"Codex + GPT-6 Luna, reasoning xhigh (ChatGPT login)","billing":"subscription","kind":"agent","done":{"privileged":408,"standard":408},"success":{"privileged":259,"standard":106},"suites":["behavior-1k","robolab","robocasa","robocasa365","robocasa-gr1","metaworldplus","vlabench","humanoidbench","kinder","robopaint","robotwin-2","mujoco-playground","dextoolbench","robowits"],"est_usd":206.21,"est_n":816,"agent_s":3089764,"agent_n":816},{"id":"codex-gpt6_astra-high","model":"GPT-6 Astra","harness":"Codex","effort":"high","route":"ChatGPT login + Azure","label":"Codex + GPT-6 Astra, reasoning high (ChatGPT login + Azure)","billing":"subscription","kind":"agent","done":{"privileged":202,"standard":291},"success":{"privileged":185,"standard":179},"suites":["behavior-1k","robolab","robocasa","robocasa365","robocasa-gr1","metaworldplus","vlabench","humanoidbench","kinder","robopaint","robotwin-2","mujoco-playground","dextoolbench","robowits"],"est_usd":4755.61,"est_n":493,"agent_s":852498,"agent_n":493},{"id":"codex-gpt6_1_sol-high","model":"GPT-6.1 Sol","harness":"Codex","effort":"high","route":"Azure + ChatGPT login","label":"Codex + GPT-6.1 Sol, reasoning high (Azure + ChatGPT login)","billing":"api","kind":"agent","done":{"privileged":138,"standard":187},"success":{"privileged":123,"standard":108},"suites":["behavior-1k","robolab","robocasa","robocasa365","robocasa-gr1","metaworldplus","vlabench","humanoidbench","kinder","robotwin-2","mujoco-playground","dextoolbench","robowits"],"est_usd":557.65,"est_n":325,"agent_s":733725,"agent_n":325},{"id":"codex-gpt6_sol-high","model":"GPT-6 Sol","harness":"Codex","effort":"high","route":"Azure + ChatGPT login","label":"Codex + GPT-6 Sol, reasoning high (Azure + ChatGPT login)","billing":"api","kind":"agent","done":{"privileged":136,"standard":180},"success":{"privileged":105,"standard":73},"suites":["behavior-1k","robolab","robocasa","robocasa365","robocasa-gr1","metaworldplus","vlabench","humanoidbench","kinder","robotwin-2","mujoco-playground","dextoolbench","robowits"],"est_usd":1069.51,"est_n":316,"agent_s":782169,"agent_n":316},{"id":"codex-gpt6_luna-xhigh-azure","model":"GPT-6 Luna","harness":"Codex","effort":"xhigh","route":"Azure API stress test","label":"Codex + GPT-6 Luna, reasoning xhigh (Azure OpenAI API, 2026-09-30 stress test)","billing":"api","kind":"agent","done":{"privileged":72,"standard":72},"success":{"privileged":47,"standard":20},"suites":["robotwin-2","mujoco-playground","dextoolbench"],"est_usd":21.36,"est_n":144,"agent_s":291483,"agent_n":144},{"id":"codex-gpt6_luna-medium-openrouter","model":"GPT-6 Luna","harness":"Codex","effort":"medium","route":"OpenRouter","label":"Codex CLI 0.157–0.159 + GPT-6 Luna, reasoning medium (OpenRouter)","billing":"openrouter","kind":"agent","done":{"privileged":21,"standard":15},"success":{"privileged":10,"standard":0},"suites":["humanoidbench","molmospaces","kinder"],"standard_reset":{"done":6,"success":2,"budget":50},"est_usd":7.42,"est_n":36,"agent_s":62119,"agent_n":36},{"id":"codex-gpt6_1_sol-medium","model":"GPT-6.1 Sol","harness":"Codex","effort":"medium","route":"OpenRouter","label":"Codex CLI 0.159.2 + GPT-6.1 Sol, reasoning medium (OpenRouter)","billing":"openrouter","kind":"agent","done":{"privileged":13,"standard":0},"success":{"privileged":6,"standard":0},"suites":["humanoidbench","kinder"],"standard_reset":{"done":13,"success":4,"budget":50},"est_usd":27.8,"est_n":13,"agent_s":37521,"agent_n":13},{"id":"codex-gpt6_luna-medium","model":"GPT-6 Luna","harness":"Codex","effort":"medium","route":"ChatGPT login","label":"Codex + GPT-6 Luna, reasoning medium (ChatGPT login)","billing":"subscription","kind":"agent","done":{"privileged":21,"standard":22},"success":{"privileged":11,"standard":3},"suites":["robotwin-2"],"est_usd":1.9,"est_n":43,"agent_s":32168,"agent_n":43},{"id":"claude_code-claude_opus5_5-medium","model":"Claude Opus 5.5","harness":"Claude Code","effort":"medium","route":"OpenRouter","label":"Claude Code 2.1.283 + Claude Opus 5.5, reasoning medium (OpenRouter)","billing":"openrouter","kind":"agent","done":{"privileged":3,"standard":0},"success":{"privileged":2,"standard":0},"suites":["humanoidbench"],"standard_reset":{"done":3,"success":0,"budget":50},"est_usd":7.1,"est_n":3,"agent_s":4625,"agent_n":3},{"id":"codex-gpt6_sol-medium","model":"GPT-6 Sol","harness":"Codex","effort":"medium","route":"OpenRouter","label":"Codex CLI 0.157.0 + GPT-6 Sol, reasoning medium (OpenRouter)","billing":"openrouter","kind":"agent","done":{"privileged":1,"standard":0},"success":{"privileged":0,"standard":0},"suites":["humanoidbench"],"standard_reset":{"done":3,"success":0,"budget":50},"est_usd":4.74,"est_n":1,"agent_s":3506,"agent_n":1}],"leaderboard":{"min_suites":10,"note":"Preliminary: runs from 23 September to 5 October 2026, some before every question moved to protocol v1.0.","notes":[{"icon":"shield","title":"Solved means marked 1","visual":"states","text":"The verifier marked the answer 1 on a fresh replay."},{"icon":"pair","title":"Two modes, never mixed","visual":"modes","text":"Privileged is the open-book mode, standard the closed-book one; they are scored apart. Standard attempts that allowed resets are not counted."},{"icon":"venn","title":"Head to head","visual":"venn","text":"Only the questions every ranked model sat count. GPT-6 Luna sat all of them; the others sat its misses plus a sample."},{"icon":"podium","title":"Ranked from {min_suites} suites","text":"A model is ranked once it has marked answers in {min_suites} suites. Smaller runs are listed under the board."},{"icon":"clock","title":"The same budget","visual":"budget","text":"One attempt per question and mode, under the same limits for every model."},{"icon":"target","title":"The exam's questions only","text":"Counts the exam's questions at {commit}."}],"shared":{"runs":["codex-gpt6_luna-xhigh","codex-gpt6_astra-high","codex-gpt6_1_sol-high","codex-gpt6_sol-high"],"trials":{"privileged":136,"standard":180},"scores":{"codex-gpt6_luna-xhigh":{"privileged":[60,136],"standard":[18,180]},"codex-gpt6_astra-high":{"privileged":[124,136],"standard":[111,180]},"codex-gpt6_1_sol-high":{"privileged":[121,136],"standard":[106,180]},"codex-gpt6_sol-high":{"privileged":[105,136],"standard":[73,180]}},"suites":{"codex-gpt6_luna-xhigh":{"behavior-1k":{"standard":[2,11],"privileged":[4,7]},"robolab":{"standard":[2,17],"privileged":[8,17]},"robocasa":{"privileged":[6,10],"standard":[1,24]},"robocasa365":{"standard":[0,17],"privileged":[5,17]},"robocasa-gr1":{"privileged":[4,15],"standard":[0,17]},"metaworldplus":{"privileged":[9,9],"standard":[3,22]},"vlabench":{"privileged":[7,10],"standard":[4,17]},"humanoidbench":{"privileged":[0,9],"standard":[0,6]},"kinder":{"standard":[0,2],"privileged":[1,1]},"robotwin-2":{"privileged":[6,11],"standard":[2,17]},"mujoco-playground":{"privileged":[3,3],"standard":[1,4]},"dextoolbench":{"privileged":[0,17],"standard":[0,17]},"robowits":{"privileged":[7,10],"standard":[3,9]}},"codex-gpt6_astra-high":{"behavior-1k":{"standard":[10,11],"privileged":[7,7]},"robolab":{"standard":[10,17],"privileged":[15,17]},"robocasa":{"privileged":[10,10],"standard":[17,24]},"robocasa365":{"standard":[9,17],"privileged":[17,17]},"robocasa-gr1":{"privileged":[13,15],"standard":[7,17]},"metaworldplus":{"privileged":[9,9],"standard":[20,22]},"vlabench":{"privileged":[10,10],"standard":[15,17]},"humanoidbench":{"privileged":[4,9],"standard":[2,6]},"kinder":{"standard":[0,2],"privileged":[1,1]},"robotwin-2":{"privileged":[11,11],"standard":[11,17]},"mujoco-playground":{"privileged":[3,3],"standard":[4,4]},"dextoolbench":{"privileged":[14,17],"standard":[0,17]},"robowits":{"privileged":[10,10],"standard":[6,9]}},"codex-gpt6_1_sol-high":{"behavior-1k":{"standard":[10,11],"privileged":[7,7]},"robolab":{"standard":[9,17],"privileged":[16,17]},"robocasa":{"privileged":[10,10],"standard":[15,24]},"robocasa365":{"standard":[7,17],"privileged":[16,17]},"robocasa-gr1":{"privileged":[15,15],"standard":[7,17]},"metaworldplus":{"privileged":[9,9],"standard":[19,22]},"vlabench":{"privileged":[10,10],"standard":[15,17]},"humanoidbench":{"privileged":[3,9],"standard":[1,6]},"kinder":{"standard":[1,2],"privileged":[1,1]},"robotwin-2":{"privileged":[11,11],"standard":[12,17]},"mujoco-playground":{"privileged":[3,3],"standard":[3,4]},"dextoolbench":{"privileged":[10,17],"standard":[0,17]},"robowits":{"privileged":[10,10],"standard":[7,9]}},"codex-gpt6_sol-high":{"behavior-1k":{"standard":[8,11],"privileged":[6,7]},"robolab":{"standard":[8,17],"privileged":[13,17]},"robocasa":{"privileged":[10,10],"standard":[9,24]},"robocasa365":{"standard":[5,17],"privileged":[10,17]},"robocasa-gr1":{"privileged":[15,15],"standard":[2,17]},"metaworldplus":{"privileged":[9,9],"standard":[19,22]},"vlabench":{"privileged":[9,10],"standard":[13,17]},"humanoidbench":{"privileged":[3,9],"standard":[1,6]},"kinder":{"standard":[0,2],"privileged":[1,1]},"robotwin-2":{"privileged":[11,11],"standard":[3,17]},"mujoco-playground":{"privileged":[3,3],"standard":[0,4]},"dextoolbench":{"privileged":[5,17],"standard":[0,17]},"robowits":{"privileged":[10,10],"standard":[5,9]}}}}},"contribute":{"recognition":{"draft":true,"lead":"Every task names its author, and every suite credits its upstream benchmark. A proposal for the maintainers:","credits":[{"what":"A suite accepted into the exam: proposal, tasks, image and evidence","credits":15},{"what":"A family of tasks added to a suite already in the exam","credits":8},{"what":"A human review of one family (PROTOCOL §2.7)","credits":3},{"what":"A verified exploit or task bug, with its fix merged","credits":3},{"what":"A full evaluation run of the exam, logs delivered","credits":5}],"threshold":15,"threshold_note":"Contributors at or above the threshold are invited as authors; core maintainers are listed first."}},"hero":{"lead":"A Terminal-Bench for generalist coding agents in robotics.","tiles":[{"suite":"roboolympics","id":"dunk","title":"Dunk","poster":"media/roboolympics/dunk.jpg?v=03e8838d","video":"media/roboolympics/wall/dunk.mp4?v=3fb535dd"},{"suite":"roboarc","id":"interlocked_claws_05","title":"Interlocked Claws 05","poster":"media/roboarc/interlocked_claws_05.jpg?v=acc8f0c2","video":"media/roboarc/wall/interlocked_claws_05.mp4?v=cc8cc31d"},{"suite":"behavior-1k","id":"setting_the_table","title":"Setting The Table","poster":"media/behavior-1k/setting_the_table.jpg?v=16726371","video":"media/behavior-1k/wall/setting_the_table.mp4?v=8b4c87be"},{"suite":"myosuite","id":"osl_run","title":"Osl Run","poster":"media/myosuite/osl_run.jpg?v=a21eb94a","video":"media/myosuite/wall/osl_run.mp4?v=a38c6892"},{"suite":"dexjoco","id":"bimanual_hanoi","title":"Bimanual Hanoi","poster":"media/dexjoco/bimanual_hanoi.jpg?v=5dcf791d","video":"media/dexjoco/wall/bimanual_hanoi.mp4?v=82349abe"},{"suite":"dexhand","id":"book_page_turn","title":"Book Page Turn","poster":"media/dexhand/book_page_turn.jpg?v=90ec4ed3","video":"media/dexhand/book_page_turn.mp4?v=221e8900"},{"suite":"mujoco-playground","id":"aloha_single_peg_insertion","title":"Aloha Single Peg Insertion","poster":"media/mujoco-playground/aloha_single_peg_insertion.jpg?v=3a1b783b","video":"media/mujoco-playground/wall/aloha_single_peg_insertion.mp4?v=e64c3098"},{"suite":"pandadyn","id":"d09_lever_launch_catch_reload","title":"D09 Lever Launch Catch Reload","poster":"media/pandadyn/d09_lever_launch_catch_reload.jpg?v=15c1f96a","video":"media/pandadyn/wall/d09_lever_launch_catch_reload.mp4?v=0f46a204"},{"suite":"robocasa-gr1","id":"can_to_drawer_close","title":"GR1 · PnPCanToDrawerClose","poster":"media/robocasa-gr1/can_to_drawer_close.jpg?v=89bc5d0d","video":"media/robocasa-gr1/wall/can_to_drawer_close.mp4?v=d2d8eb41"},{"suite":"ambench","id":"lemon_harvesting","title":"Lemon Harvesting","poster":"media/ambench/lemon_harvesting.jpg?v=6b2281e1","video":"media/ambench/wall/lemon_harvesting.mp4?v=f31532e9"},{"suite":"robowits","id":"dominos","title":"Dominos","poster":"media/robowits/dominos.jpg?v=55ae1e4c","video":"media/robowits/dominos.mp4?v=76c71b5f"},{"suite":"robopaint","id":"oil_starrynight_i01","title":"Oil · Starrynight","poster":"media/robopaint/oil_starrynight_i01.jpg?v=cbd92a34","video":"media/robopaint/wall/oil_starrynight_i01.mp4?v=1e71c13c"},{"suite":"robotwin-2","id":"hanging_mug","title":"Hanging mug","poster":"media/robotwin-2/hanging_mug.jpg?v=3f4d4f15","video":"media/robotwin-2/wall/hanging_mug.mp4?v=572b3324"},{"suite":"dextoolbench","id":"hammer_claw_hammer_swing_side","title":"Claw hammer · swing side","poster":"media/dextoolbench/hammer_claw_hammer_swing_side.jpg?v=f2a5a3ba","video":"media/dextoolbench/wall/hammer_claw_hammer_swing_side.mp4?v=8948270f"},{"suite":"robocasa365","id":"open_dishwasher","title":"Atomic · OpenDishwasher","poster":"media/robocasa365/open_dishwasher.jpg?v=7920f46c","video":"media/robocasa365/wall/open_dishwasher.mp4?v=bfb5b95c"},{"suite":"maniguard","id":"cabinet_pickup_i22","title":"Cabinet Pickup I22","poster":"media/maniguard/cabinet_pickup_i22.jpg?v=48ae2f29","video":"media/maniguard/wall/cabinet_pickup_i22.mp4?v=88c060bf"},{"suite":"metaworldplus","id":"basketball","title":"Meta-World · Basketball","poster":"media/metaworldplus/basketball.jpg?v=c0eb7f60","video":"media/metaworldplus/wall/basketball.mp4?v=2cb3df07"},{"suite":"vlabench","id":"hammer_nail_and_hang_picture","title":"Manipulation · hammer_nail_and_hang_picture","poster":"media/vlabench/hammer_nail_and_hang_picture.jpg?v=4e30013b","video":"media/vlabench/wall/hammer_nail_and_hang_picture.mp4?v=d90f4634"},{"suite":"yamlab","id":"hang_mug_on_tree_i10","title":"Hang Mug On Tree I10","poster":"media/yamlab/hang_mug_on_tree_i10.jpg?v=5d1d9093","video":"media/yamlab/hang_mug_on_tree_i10.mp4?v=407ba130"},{"suite":"robolab","id":"reorient_red_mug","title":"Reorient Red Mug","poster":"media/robolab/reorient_red_mug.jpg?v=ed39554f","video":"media/robolab/reorient_red_mug.mp4?v=0570d703"},{"suite":"robodojo","id":"match_and_pick_from_conveyor","title":"Match And Pick From Conveyor","poster":"media/robodojo/match_and_pick_from_conveyor.jpg?v=51b48f49","video":"media/robodojo/wall/match_and_pick_from_conveyor.mp4?v=bf543ed4"},{"suite":"roboarc","id":"knot_untying_03","title":"Knot Untying 03","poster":"media/roboarc/knot_untying_03.jpg?v=9eab5621","video":"media/roboarc/knot_untying_03.mp4?v=fbeae54b"},{"suite":"behavior-1k","id":"cook_brussels_sprouts","title":"Cook Brussels Sprouts","poster":"media/behavior-1k/cook_brussels_sprouts.jpg?v=3817af1a","video":"media/behavior-1k/wall/cook_brussels_sprouts.mp4?v=812bd85d"},{"suite":"myosuite","id":"baoding","title":"Baoding","poster":"media/myosuite/baoding.jpg?v=3d79e192","video":"media/myosuite/wall/baoding.mp4?v=404f0ffc"},{"suite":"dexjoco","id":"water_plant","title":"Water Plant","poster":"media/dexjoco/water_plant.jpg?v=389861dd","video":"media/dexjoco/wall/water_plant.mp4?v=9112acb7"},{"suite":"dexhand","id":"jar_lid_open","title":"Jar Lid Open","poster":"media/dexhand/jar_lid_open.jpg?v=67d26ee7","video":"media/dexhand/jar_lid_open.mp4?v=6580bd30"},{"suite":"pandadyn","id":"s09_toss_recatch_delivery","title":"S09 Toss Recatch Delivery","poster":"media/pandadyn/s09_toss_recatch_delivery.jpg?v=3a6d9b52","video":"media/pandadyn/s09_toss_recatch_delivery.mp4?v=19ff64cc"},{"suite":"robocasa-gr1","id":"milk_to_microwave_close","title":"GR1 · PnPMilkToMicrowaveClose","poster":"media/robocasa-gr1/milk_to_microwave_close.jpg?v=35bf02b1","video":"media/robocasa-gr1/milk_to_microwave_close.mp4?v=cec4d20b"},{"suite":"ambench","id":"open_door","title":"Open Door","poster":"media/ambench/open_door.jpg?v=9ac5dc81","video":"media/ambench/open_door.mp4?v=14043aa2"},{"suite":"robowits","id":"ball_into_bottle","title":"Ball Into Bottle","poster":"media/robowits/ball_into_bottle.jpg?v=2237b953","video":"media/robowits/wall/ball_into_bottle.mp4?v=3e4bddd9"},{"suite":"robopaint","id":"acrylic_greatwave","title":"Acrylic · Greatwave","poster":"media/robopaint/acrylic_greatwave.jpg?v=97fad004","video":"media/robopaint/acrylic_greatwave.mp4?v=f7444716"},{"suite":"robocasa365","id":"close_stand_mixer_head","title":"Atomic · CloseStandMixerHead","poster":"media/robocasa365/close_stand_mixer_head.jpg?v=1e675e4e","video":"media/robocasa365/wall/close_stand_mixer_head.mp4?v=7b90f795"},{"suite":"maniguard","id":"jar_transport_i02","title":"Jar Transport I02","poster":"media/maniguard/jar_transport_i02.jpg?v=54107a29","video":"media/maniguard/wall/jar_transport_i02.mp4?v=3ab937d6"},{"suite":"vlabench","id":"texas_holdem","title":"Manipulation · texas_holdem","poster":"media/vlabench/texas_holdem.jpg?v=30594a6d","video":"media/vlabench/wall/texas_holdem.mp4?v=e1d84a0d"},{"suite":"yamlab","id":"put_pot_on_cooktop_i03","title":"Put Pot On Cooktop I03","poster":"media/yamlab/put_pot_on_cooktop_i03.jpg?v=4b0c19e0","video":"media/yamlab/put_pot_on_cooktop_i03.mp4?v=182f0621"},{"suite":"robolab","id":"block_stacking_specified_order","title":"Block Stacking Specified Order","poster":"media/robolab/block_stacking_specified_order.jpg?v=3b9b6d5b","video":"media/robolab/wall/block_stacking_specified_order.mp4?v=b43cd093"},{"suite":"robodojo","id":"deposit_coin","title":"Deposit Coin","poster":"media/robodojo/deposit_coin.jpg?v=1cc7029a","video":"media/robodojo/wall/deposit_coin.mp4?v=6cdc348a"},{"suite":"roboarc","id":"buzz_wire_3d_06","title":"Buzz Wire 3D 06","poster":"media/roboarc/buzz_wire_3d_06.jpg?v=3148b43a","video":"media/roboarc/buzz_wire_3d_06.mp4?v=dc72286f"},{"suite":"behavior-1k","id":"putting_up_Christmas_decorations_inside","title":"Putting Up Christmas Decorations Inside","poster":"media/behavior-1k/putting_up_Christmas_decorations_inside.jpg?v=b543c98e","video":"media/behavior-1k/putting_up_Christmas_decorations_inside.mp4?v=837ebaa7"},{"suite":"myosuite","id":"soccer","title":"Soccer","poster":"media/myosuite/soccer.jpg?v=41117f30","video":"media/myosuite/wall/soccer.mp4?v=6119cf10"},{"suite":"dexjoco","id":"bimanual_photograph","title":"Bimanual Photograph","poster":"media/dexjoco/bimanual_photograph.jpg?v=e613a880","video":"media/dexjoco/wall/bimanual_photograph.mp4?v=95e362f2"},{"suite":"dexhand","id":"essential_pen_regrasp","title":"Essential Pen Regrasp","poster":"media/dexhand/essential_pen_regrasp.jpg?v=a395791b","video":"media/dexhand/essential_pen_regrasp.mp4?v=7315574b"},{"suite":"ambench","id":"ndt","title":"Ndt","poster":"media/ambench/ndt.jpg?v=33d0d1f0","video":"media/ambench/ndt.mp4?v=b107f8ff"},{"suite":"robowits","id":"stand_bulb","title":"Stand Bulb","poster":"media/robowits/stand_bulb.jpg?v=6c6fffcc","video":"media/robowits/stand_bulb.mp4?v=4d6c8638"},{"suite":"robopaint","id":"caoshu_zhongtiangechuifen","title":"Caoshu · Zhongtiangechuifen","poster":"media/robopaint/caoshu_zhongtiangechuifen.jpg?v=04fcfeab","video":"media/robopaint/caoshu_zhongtiangechuifen.mp4?v=9fa46549"},{"suite":"robocasa365","id":"oven_broil_fish","title":"Composite · OvenBroilFish","poster":"media/robocasa365/oven_broil_fish.jpg?v=8f576e9e","video":"media/robocasa365/wall/oven_broil_fish.mp4?v=29bff6be"},{"suite":"maniguard","id":"lid_transport_i07","title":"Lid Transport I07","poster":"media/maniguard/lid_transport_i07.jpg?v=0397e229","video":"media/maniguard/wall/lid_transport_i07.mp4?v=8dd1a894"},{"suite":"vlabench","id":"insert_bloom_flower","title":"Manipulation · insert_bloom_flower","poster":"media/vlabench/insert_bloom_flower.jpg?v=c7a88728","video":"media/vlabench/insert_bloom_flower.mp4?v=fb9459f3"},{"suite":"robolab","id":"take_mugs_off_of_shelf","title":"Take Mugs Off Of Shelf","poster":"media/robolab/take_mugs_off_of_shelf.jpg?v=ed5e70f1","video":"media/robolab/take_mugs_off_of_shelf.mp4?v=b79e4dfc"},{"suite":"roboarc","id":"blanket_fold_03","title":"Blanket Fold 03","poster":"media/roboarc/blanket_fold_03.jpg?v=2c55369f","video":"media/roboarc/blanket_fold_03.mp4?v=1f27f958"},{"suite":"behavior-1k","id":"chopping_wood","title":"Chopping Wood","poster":"media/behavior-1k/chopping_wood.jpg?v=4d536eca","video":"media/behavior-1k/chopping_wood.mp4?v=bd04b4bf"},{"suite":"myosuite","id":"table_tennis","title":"Table Tennis","poster":"media/myosuite/table_tennis.jpg?v=d3b2c321","video":"media/myosuite/wall/table_tennis.mp4?v=190ddc2d"},{"suite":"dexjoco","id":"fold_glasses","title":"Fold Glasses","poster":"media/dexjoco/fold_glasses.jpg?v=fdd9b66e","video":"media/dexjoco/wall/fold_glasses.mp4?v=b7762cb5"},{"suite":"dexhand","id":"bicycle_pump_stroke","title":"Bicycle Pump Stroke","poster":"media/dexhand/bicycle_pump_stroke.jpg?v=a3915e52","video":"media/dexhand/bicycle_pump_stroke.mp4?v=58246322"},{"suite":"robowits","id":"stand_pages","title":"Stand Pages","poster":"media/robowits/stand_pages.jpg?v=4ba023b1","video":"media/robowits/wall/stand_pages.mp4?v=a0bfc1b8"},{"suite":"robopaint","id":"oil_monalisa_i01","title":"Oil · Monalisa","poster":"media/robopaint/oil_monalisa_i01.jpg?v=6393e7ae","video":"media/robopaint/oil_monalisa_i01.mp4?v=65572639"},{"suite":"robolab","id":"unstack_rubiks_cube","title":"Unstack Rubiks Cube","poster":"media/robolab/unstack_rubiks_cube.jpg?v=f9dbd182","video":"media/robolab/wall/unstack_rubiks_cube.mp4?v=15adc5a3"},{"suite":"roboarc","id":"lock_picking_02","title":"Lock Picking 02","poster":"media/roboarc/lock_picking_02.jpg?v=c3d36dbc","video":"media/roboarc/lock_picking_02.mp4?v=2f0501df"},{"suite":"behavior-1k","id":"hanging_pictures","title":"Hanging Pictures","poster":"media/behavior-1k/hanging_pictures.jpg?v=a19c12cc","video":"media/behavior-1k/hanging_pictures.mp4?v=203a88e7"},{"suite":"myosuite","id":"die_reorient","title":"Die Reorient","poster":"media/myosuite/die_reorient.jpg?v=2a36978c","video":"media/myosuite/wall/die_reorient.mp4?v=4756e655"},{"suite":"robowits","id":"hold_cup","title":"Hold Cup","poster":"media/robowits/hold_cup.jpg?v=8a1f5220","video":"media/robowits/wall/hold_cup.mp4?v=350a16c2"},{"suite":"robopaint","id":"kaishu_yongzibafa","title":"Kaishu · Yongzibafa","poster":"media/robopaint/kaishu_yongzibafa.jpg?v=1ca9fd98","video":"media/robopaint/kaishu_yongzibafa.mp4?v=55ebb829"},{"suite":"roboarc","id":"rope_threading_04","title":"Rope Threading 04","poster":"media/roboarc/rope_threading_04.jpg?v=c07283cd","video":"media/roboarc/rope_threading_04.mp4?v=3ad3c101"},{"suite":"behavior-1k","id":"carrying_out_garden_furniture","title":"Carrying Out Garden Furniture","poster":"media/behavior-1k/carrying_out_garden_furniture.jpg?v=a7e5cca3","video":"media/behavior-1k/carrying_out_garden_furniture.mp4?v=b5acff1e"},{"suite":"myosuite","id":"pen_twirl","title":"Pen Twirl","poster":"media/myosuite/pen_twirl.jpg?v=97ad8e2a","video":"media/myosuite/wall/pen_twirl.mp4?v=73876f8b"},{"suite":"robowits","id":"raise_platform","title":"Raise Platform","poster":"media/robowits/raise_platform.jpg?v=8fa51128","video":"media/robowits/wall/raise_platform.mp4?v=54919c49"},{"suite":"roboarc","id":"hedgehog_in_cage_05","title":"Hedgehog In Cage 05","poster":"media/roboarc/hedgehog_in_cage_05.jpg?v=a669931c","video":"media/roboarc/hedgehog_in_cage_05.mp4?v=24fdacee"},{"suite":"behavior-1k","id":"rearranging_kitchen_furniture","title":"Rearranging Kitchen Furniture","poster":"media/behavior-1k/rearranging_kitchen_furniture.jpg?v=4e2ccb9a","video":"media/behavior-1k/rearranging_kitchen_furniture.mp4?v=27e2c3f3"},{"suite":"roboarc","id":"fiber_bundling_06","title":"Fiber Bundling 06","poster":"media/roboarc/fiber_bundling_06.jpg?v=ffc3574a","video":"media/roboarc/fiber_bundling_06.mp4?v=83ce49a7"},{"suite":"behavior-1k","id":"make_pizza","title":"Make Pizza","poster":"media/behavior-1k/make_pizza.jpg?v=a1bd371b","video":"media/behavior-1k/wall/make_pizza.mp4?v=dc3d0c59"},{"suite":"roboarc","id":"cast_keyhole_03","title":"Cast Keyhole 03","poster":"media/roboarc/cast_keyhole_03.jpg?v=723e3c90","video":"media/roboarc/cast_keyhole_03.mp4?v=bfe9ac9d"},{"suite":"roboarc","id":"horseshoe_ring_02","title":"Horseshoe Ring 02","poster":"media/roboarc/horseshoe_ring_02.jpg?v=81b8ea9a","video":"media/roboarc/horseshoe_ring_02.mp4?v=734e7bf9"},{"suite":"roboarc","id":"slotted_post_ring_03","title":"Slotted Post Ring 03","poster":"media/roboarc/slotted_post_ring_03.jpg?v=90357263","video":"media/roboarc/wall/slotted_post_ring_03.mp4?v=793fe426"},{"suite":"roboarc","id":"cast_coil_04","title":"Cast Coil 04","poster":"media/roboarc/cast_coil_04.jpg?v=4cb48a02","video":"media/roboarc/wall/cast_coil_04.mp4?v=9dbe55e0"},{"suite":"roboarc","id":"blanket_flatten_04","title":"Blanket Flatten 04","poster":"media/roboarc/blanket_flatten_04.jpg?v=d3b1c0ef","video":"media/roboarc/wall/blanket_flatten_04.mp4?v=71ed4297"},{"suite":"roboarc","id":"interlocked_claws_09","title":"Interlocked Claws 09","poster":"media/roboarc/interlocked_claws_09.jpg?v=808ae992","video":"media/roboarc/interlocked_claws_09.mp4?v=4ab51fc8"},{"suite":"roboarc","id":"knot_untying_05","title":"Knot Untying 05","poster":"media/roboarc/knot_untying_05.jpg?v=57549292","video":"media/roboarc/wall/knot_untying_05.mp4?v=71acd7d9"},{"suite":"roboarc","id":"rope_threading_07","title":"Rope Threading 07","poster":"media/roboarc/rope_threading_07.jpg?v=acbf3f35","video":"media/roboarc/wall/rope_threading_07.mp4?v=0cd9e86d"},{"suite":"roboarc","id":"buzz_wire_3d_09","title":"Buzz Wire 3D 09","poster":"media/roboarc/buzz_wire_3d_09.jpg?v=01307a25","video":"media/roboarc/wall/buzz_wire_3d_09.mp4?v=2f9a4f44"},{"suite":"roboarc","id":"lock_picking_04","title":"Lock Picking 04","poster":"media/roboarc/lock_picking_04.jpg?v=cbbab161","video":"media/roboarc/wall/lock_picking_04.mp4?v=2e0c228a"},{"suite":"roboarc","id":"hedgehog_in_cage_08","title":"Hedgehog In Cage 08","poster":"media/roboarc/hedgehog_in_cage_08.jpg?v=4f3fae1f","video":"media/roboarc/wall/hedgehog_in_cage_08.mp4?v=3c063718"},{"suite":"roboarc","id":"fiber_bundling_09","title":"Fiber Bundling 09","poster":"media/roboarc/fiber_bundling_09.jpg?v=d910ed54","video":"media/roboarc/wall/fiber_bundling_09.mp4?v=c82684d6"},{"suite":"roboarc","id":"lack","title":"Lack","poster":"media/ikea-mujoco/lack.jpg?v=2d957909","video":"media/ikea-mujoco/lack.mp4?v=1d42c53c"},{"suite":"roboarc","id":"eket","title":"Eket","poster":"media/ikea-mujoco/eket.jpg?v=5d524e83","video":"media/ikea-mujoco/wall/eket.mp4?v=501de5c1"},{"suite":"roboarc","id":"svallet","title":"Svallet","poster":"media/ikea-mujoco/svallet.jpg?v=ad182e72","video":"media/ikea-mujoco/svallet.mp4?v=846d2a2c"},{"suite":"roboarc","id":"variera","title":"Variera","poster":"media/ikea-mujoco/variera.jpg?v=6d4c6343","video":"media/ikea-mujoco/wall/variera.mp4?v=bfb3335e"},{"suite":"roboarc","id":"vesken","title":"Vesken","poster":"media/ikea-mujoco/vesken.jpg?v=d182d0ac","video":"media/ikea-mujoco/wall/vesken.mp4?v=29fd0fbb"}],"front":["roboarc","myosuite","roboolympics","behavior-1k","dexjoco","robocasa365","dexhand","pandadyn","robopaint","robotwin-2","ambench","robowits","robocasa-gr1","maniguard","yamlab","mujoco-playground","dextoolbench"],"back":["metaworldplus","robodojo","robolab","vlabench"],"centre":"roboarc/interlocked_claws_05"},"taxmap":{"rows":[{"id":"perception-understanding","name":"Perception & Understanding","description":"Perception, object/state recognition, semantic grounding, spatial understanding.","n":179},{"id":"planning-reasoning","name":"Planning & Reasoning","description":"Task planning, sequencing, tool/affordance reasoning, geometric reasoning.","n":291},{"id":"control-coordination","name":"Control & Coordination","description":"Manipulation/control, contact-rich execution, locomotion control, multi-arm or whole-body coordination.","n":469},{"id":"feedback-adaptation","name":"Feedback & Adaptation","description":"Closed-loop correction, failure recovery, online adaptation, exploration based on feedback.","n":61}],"cols":[{"id":"manipulation","name":"Manipulation","description":"Fixed-base object manipulation, grasping, pushing, insertion, tool use, articulated/deformable object interaction.","n":332},{"id":"locomotion-stability","name":"Locomotion & Stability","description":"Walking, running, balancing, terrain traversal, dynamic stabilization.","n":4},{"id":"navigation-exploration","name":"Navigation & Exploration","description":"Navigation, target search, exploration, spatial memory.","n":40},{"id":"mobile-whole-body-manipulation","name":"Mobile / Whole-body Manipulation","description":"Tasks combining navigation/locomotion with manipulation or requiring coordinated whole-body motion.","n":132},{"id":"interaction-collaboration","name":"Interaction & Collaboration","description":"Human-robot interaction, multi-robot collaboration, handover, cooperative tasks.","n":3}],"cells":{"perception-understanding|manipulation":{"n":112,"suites":8,"ex":[{"s":"robolab","id":"apple_and_yogurt_in_bowl","t":"Apple And Yogurt In Bowl","p":"media/robolab/apple_and_yogurt_in_bowl.jpg?v=5e6ca2b7"},{"s":"robocasa","id":"restock_pantry","t":"Multi-stage · RestockPantry","p":"media/robocasa/restock_pantry.jpg?v=2d50ad0c"},{"s":"robocasa365","id":"choose_measuring_cup","t":"Composite · ChooseMeasuringCup","p":"media/robocasa365/choose_measuring_cup.jpg?v=1e508ca8"},{"s":"vlabench","id":"select_billiards_common_sense","t":"Common sense · select_billiards_common_sense","p":"media/vlabench/select_billiards_common_sense.jpg?v=d4df171d"}]},"perception-understanding|navigation-exploration":{"n":28,"suites":3,"ex":[{"s":"behavior-1k","id":"boxing_food_after_dinner","t":"Boxing Food After Dinner","p":"media/behavior-1k/boxing_food_after_dinner.jpg?v=781ec0d6"},{"s":"molmospaces","id":"navigate_to","t":"RB-Y1 · Navigate to","p":"media/molmospaces/navigate_to.jpg?v=9f30c208"},{"s":"kinder","id":"dynamo","t":"Dynamic3D · Dynamo","p":"media/kinder/dynamo.jpg?v=0af57458"}]},"perception-understanding|mobile-whole-body-manipulation":{"n":66,"suites":3,"ex":[{"s":"behavior-1k","id":"assembling_gift_baskets","t":"Assembling Gift Baskets","p":"media/behavior-1k/assembling_gift_baskets.jpg?v=0f0ce2c1"},{"s":"robocasa365","id":"beverage_sorting","t":"Composite · BeverageSorting","p":"media/robocasa365/beverage_sorting.jpg?v=b183514b"},{"s":"kinder","id":"dynamo","t":"Dynamic3D · Dynamo","p":"media/kinder/dynamo.jpg?v=0af57458"}]},"perception-understanding|interaction-collaboration":{"n":1,"suites":1,"ex":[{"s":"robotwin-2","id":"handover_block","t":"Handover block","p":"media/robotwin-2/handover_block.jpg?v=7cb81fdd"}]},"planning-reasoning|manipulation":{"n":165,"suites":11,"ex":[{"s":"robolab","id":"apple_and_yogurt_in_bowl","t":"Apple And Yogurt In Bowl","p":"media/robolab/apple_and_yogurt_in_bowl.jpg?v=5e6ca2b7"},{"s":"robocasa","id":"arrange_vegetables","t":"Multi-stage · ArrangeVegetables","p":"media/robocasa/arrange_vegetables.jpg?v=fcaa2980"},{"s":"robocasa365","id":"bread_and_cheese","t":"Composite · BreadAndCheese","p":"media/robocasa365/bread_and_cheese.jpg?v=82d73895"},{"s":"robocasa-gr1","id":"bottle_to_cabinet_close","t":"GR1 · PnPBottleToCabinetClose","p":"media/robocasa-gr1/bottle_to_cabinet_close.jpg?v=8af0dc71"}]},"planning-reasoning|navigation-exploration":{"n":38,"suites":3,"ex":[{"s":"behavior-1k","id":"boxing_food_after_dinner","t":"Boxing Food After Dinner","p":"media/behavior-1k/boxing_food_after_dinner.jpg?v=781ec0d6"},{"s":"humanoidbench","id":"manip_door","t":"Manipulation · Door","p":"media/humanoidbench/manip_door.jpg?v=6634feb3"},{"s":"kinder","id":"dynamo","t":"Dynamic3D · Dynamo","p":"media/kinder/dynamo.jpg?v=0af57458"}]},"planning-reasoning|mobile-whole-body-manipulation":{"n":126,"suites":4,"ex":[{"s":"behavior-1k","id":"assembling_gift_baskets","t":"Assembling Gift Baskets","p":"media/behavior-1k/assembling_gift_baskets.jpg?v=0f0ce2c1"},{"s":"robocasa365","id":"add_lemon_to_fish","t":"Composite · AddLemonToFish","p":"media/robocasa365/add_lemon_to_fish.jpg?v=3bd8666b"},{"s":"humanoidbench","id":"manip_bookshelf_simple","t":"Manipulation · Bookshelf simple","p":"media/humanoidbench/manip_bookshelf_simple.jpg?v=94c3aee9"},{"s":"kinder","id":"constrained_cupboard","t":"Dynamic3D · Constrained cupboard","p":"media/kinder/constrained_cupboard.jpg?v=63a002c4"}]},"planning-reasoning|interaction-collaboration":{"n":3,"suites":2,"ex":[{"s":"robotwin-2","id":"handover_block","t":"Handover block","p":"media/robotwin-2/handover_block.jpg?v=7cb81fdd"},{"s":"mujoco-playground","id":"aloha_hand_over","t":"Aloha Hand Over","p":"media/mujoco-playground/aloha_hand_over.jpg?v=286f8599"}]},"control-coordination|manipulation":{"n":332,"suites":12,"ex":[{"s":"robolab","id":"apple_and_yogurt_in_bowl","t":"Apple And Yogurt In Bowl","p":"media/robolab/apple_and_yogurt_in_bowl.jpg?v=5e6ca2b7"},{"s":"robocasa","id":"arrange_vegetables","t":"Multi-stage · ArrangeVegetables","p":"media/robocasa/arrange_vegetables.jpg?v=fcaa2980"},{"s":"robocasa365","id":"close_drawer","t":"Atomic · CloseDrawer","p":"media/robocasa365/close_drawer.jpg?v=b02bfa3c"},{"s":"robocasa-gr1","id":"bottle_to_cabinet_close","t":"GR1 · PnPBottleToCabinetClose","p":"media/robocasa-gr1/bottle_to_cabinet_close.jpg?v=8af0dc71"}]},"control-coordination|locomotion-stability":{"n":4,"suites":1,"ex":[{"s":"humanoidbench","id":"loco_balance_hard","t":"Locomotion · Balance hard","p":"media/humanoidbench/loco_balance_hard.jpg?v=97a324f4"}]},"control-coordination|navigation-exploration":{"n":40,"suites":4,"ex":[{"s":"behavior-1k","id":"boxing_food_after_dinner","t":"Boxing Food After Dinner","p":"media/behavior-1k/boxing_food_after_dinner.jpg?v=781ec0d6"},{"s":"humanoidbench","id":"manip_door","t":"Manipulation · Door","p":"media/humanoidbench/manip_door.jpg?v=6634feb3"},{"s":"molmospaces","id":"navigate_to","t":"RB-Y1 · Navigate to","p":"media/molmospaces/navigate_to.jpg?v=9f30c208"},{"s":"kinder","id":"dynamo","t":"Dynamic3D · Dynamo","p":"media/kinder/dynamo.jpg?v=0af57458"}]},"control-coordination|mobile-whole-body-manipulation":{"n":132,"suites":5,"ex":[{"s":"behavior-1k","id":"assembling_gift_baskets","t":"Assembling Gift Baskets","p":"media/behavior-1k/assembling_gift_baskets.jpg?v=0f0ce2c1"},{"s":"robocasa365","id":"add_lemon_to_fish","t":"Composite · AddLemonToFish","p":"media/robocasa365/add_lemon_to_fish.jpg?v=3bd8666b"},{"s":"humanoidbench","id":"manip_bookshelf_simple","t":"Manipulation · Bookshelf simple","p":"media/humanoidbench/manip_bookshelf_simple.jpg?v=94c3aee9"},{"s":"molmospaces","id":"open_door","t":"RB-Y1 · Open door","p":"media/molmospaces/open_door.jpg?v=a37f562c"}]},"control-coordination|interaction-collaboration":{"n":3,"suites":2,"ex":[{"s":"robotwin-2","id":"handover_block","t":"Handover block","p":"media/robotwin-2/handover_block.jpg?v=7cb81fdd"},{"s":"mujoco-playground","id":"aloha_hand_over","t":"Aloha Hand Over","p":"media/mujoco-playground/aloha_hand_over.jpg?v=286f8599"}]},"feedback-adaptation|manipulation":{"n":3,"suites":3,"ex":[{"s":"robolab","id":"cubes_and_blocks_in_bin","t":"Cubes And Blocks In Bin","p":"media/robolab/cubes_and_blocks_in_bin.jpg?v=a9dde171"},{"s":"vlabench","id":"find_unseen_object","t":"Manipulation · find_unseen_object","p":"media/vlabench/find_unseen_object.jpg?v=ddf71721"},{"s":"robotwin-2","id":"put_bottles_dustbin","t":"Put bottles dustbin","p":"media/robotwin-2/put_bottles_dustbin.jpg?v=5fe48754"}]},"feedback-adaptation|locomotion-stability":{"n":2,"suites":1,"ex":[{"s":"humanoidbench","id":"loco_balance_hard","t":"Locomotion · Balance hard","p":"media/humanoidbench/loco_balance_hard.jpg?v=97a324f4"}]},"feedback-adaptation|navigation-exploration":{"n":24,"suites":3,"ex":[{"s":"behavior-1k","id":"boxing_food_after_dinner","t":"Boxing Food After Dinner","p":"media/behavior-1k/boxing_food_after_dinner.jpg?v=781ec0d6"},{"s":"molmospaces","id":"navigate_to","t":"RB-Y1 · Navigate to","p":"media/molmospaces/navigate_to.jpg?v=9f30c208"},{"s":"kinder","id":"dynamo","t":"Dynamic3D · Dynamo","p":"media/kinder/dynamo.jpg?v=0af57458"}]},"feedback-adaptation|mobile-whole-body-manipulation":{"n":55,"suites":2,"ex":[{"s":"behavior-1k","id":"assembling_gift_baskets","t":"Assembling Gift Baskets","p":"media/behavior-1k/assembling_gift_baskets.jpg?v=0f0ce2c1"},{"s":"kinder","id":"dynamo","t":"Dynamic3D · Dynamo","p":"media/kinder/dynamo.jpg?v=0af57458"}]}},"total":469,"per_question":{"1":132,"2":193,"3":94,"4":50},"top_mix":{"ids":["planning-reasoning","control-coordination"],"n":149}}}