{ "suite": "level_2", "step": 80000, "num_tasks": 28, "tasks_evaluated": 28, "evaluation_complete": true, "pending_tasks": [], "episodes_per_task": 50, "episode_seeds": "1000-1049", "mean_success_rate": 0.2414285714285714, "per_task": { "HangMug-v1": { "success_rate": 0.42, "successes": 21, "return_mean": 19.1683, "embodiment": "rigid_body/single_arm_gripper" }, "PivotBook-v1": { "success_rate": 0.16, "successes": 8, "return_mean": 179.89354, "embodiment": "rigid_body/single_arm_gripper" }, "OpenBoxRetrieval-v1": { "success_rate": 0.04, "successes": 2, "return_mean": 187.55988000000002, "embodiment": "rigid_body/single_arm_gripper" }, "KitAssembly-v1": { "success_rate": 0.0, "successes": 0, "return_mean": 192.14, "embodiment": "rigid_body/single_arm_gripper" }, "GraspTwoBlocks-v1": { "success_rate": 0.56, "successes": 28, "return_mean": 226.9707, "embodiment": "rigid_body/single_arm_dex" }, "ReorientPen-v1": { "success_rate": 0.3, "successes": 15, "return_mean": 111.66706, "embodiment": "rigid_body/single_arm_dex" }, "DexOpenDoor-v1": { "success_rate": 0.28, "successes": 14, "return_mean": 325.17524000000003, "embodiment": "rigid_body/single_arm_dex" }, "HammerNailOut-v1": { "success_rate": 0.32, "successes": 16, "return_mean": -24.495720000000002, "embodiment": "rigid_body/single_arm_dex" }, "DualArmLiftBall-v1": { "success_rate": 0.36, "successes": 18, "return_mean": 30.12464, "embodiment": "rigid_body/dual_arm_gripper" }, "DualArmBookPickPlace-v1": { "success_rate": 0.0, "successes": 0, "return_mean": 575.62944, "embodiment": "rigid_body/dual_arm_gripper" }, "DualArmPegInHole-v1": { "success_rate": 0.18, "successes": 9, "return_mean": 62.280640000000005, "embodiment": "rigid_body/dual_arm_gripper" }, "DualArmCleanupPotEnv-v1": { "success_rate": 0.0, "successes": 0, "return_mean": 99.78744, "embodiment": "rigid_body/dual_arm_gripper" }, "DualArmOpenLaptop-v1": { "success_rate": 0.34, "successes": 17, "return_mean": 1389.98766, "embodiment": "rigid_body/dual_arm_dex" }, "DualArmThrowBlock-v1": { "success_rate": 0.2, "successes": 10, "return_mean": 2159.683, "embodiment": "rigid_body/dual_arm_dex" }, "DualArmJugglingBall-v1": { "success_rate": 0.0, "successes": 0, "return_mean": -5060.75764, "embodiment": "rigid_body/dual_arm_dex" }, "DualArmOpenBottle-v1": { "success_rate": 0.0, "successes": 0, "return_mean": 2124.39504, "embodiment": "rigid_body/dual_arm_dex" }, "DualArmPlaceRopeOverRod-v1": { "success_rate": 0.48, "successes": 24, "return_mean": 332.42735999999996, "embodiment": "deformable/dual_arm_dex" }, "DualArmStuffCloth-v1": { "success_rate": 0.44, "successes": 22, "return_mean": 34.17728, "embodiment": "deformable/dual_arm_dex" }, "DualArmPouringTwoCups-v1": { "success_rate": 0.0, "successes": 0, "return_mean": 1920.35364, "embodiment": "deformable/dual_arm_dex" }, "DualArmClothAlign-v1": { "success_rate": 0.5, "successes": 25, "return_mean": 27.72722, "embodiment": "deformable/dual_arm_gripper" }, "DualArmPushObject-v1": { "success_rate": 0.0, "successes": 0, "return_mean": 0.04186, "embodiment": "deformable/dual_arm_gripper" }, "DualArmStraightenRope-v1": { "success_rate": 0.24, "successes": 12, "return_mean": 77.2212, "embodiment": "deformable/dual_arm_gripper" }, "ClothWipe-v1": { "success_rate": 0.24, "successes": 12, "return_mean": 290.20684, "embodiment": "deformable/single_arm_dex" }, "RollingPin-v1": { "success_rate": 0.38, "successes": 19, "return_mean": 245.92406, "embodiment": "deformable/single_arm_dex" }, "RopeBridgeDex-v1": { "success_rate": 0.32, "successes": 16, "return_mean": 314.65782, "embodiment": "deformable/single_arm_dex" }, "FoldShirt-v1": { "success_rate": 0.62, "successes": 31, "return_mean": 189.38598, "embodiment": "deformable/single_arm_gripper" }, "ScoopWater-v1": { "success_rate": 0.26, "successes": 13, "return_mean": -49.51626, "embodiment": "deformable/single_arm_gripper" }, "StraightenTwoRopesSequential-v1": { "success_rate": 0.12, "successes": 6, "return_mean": 2140.54834, "embodiment": "deformable/single_arm_gripper" } }, "note": "Only the final checkpoint (step 80000) was evaluated, 50 episodes per task; per-episode rows are in eval_reports/.json.", "campaign": "pi05_levels_20261028" }