{ "dataset_repo_id": "rslxcvg/banana_act_direct_color_simple_v1_molmo_compat", "dataset_root": "/mnt/vla_picknplace/outputs/lerobot/banana_act_direct_color_simple_v1_molmo_compat", "policy_path": "/mnt/vla_picknplace/outputs/molmoact2/molmoact2_overnight_frombase_prod_r128_c010_rw3_w8_gpu5_20260519_fullcoverage_v1/checkpoints/010000/pretrained_model", "decision_manifest": "/mnt/vla_picknplace/outputs/molmoact2/decision_manifests/banana_act_direct_color_simple_v1_molmo_compat_manifest.parquet", "dry_run": false, "sample_selection": "branch_onset", "num_samples": 12, "num_inference_steps": 10, "decision_action_index": 5, "prompt_templates": { "red": "Put the banana in the red bowl.", "green": "Put the banana in the green bowl.", "blue": "Put the banana in the blue bowl." }, "shoulder_pan_range_mean": 18.936976432800293, "shoulder_pan_range_median": 18.53101348876953, "shoulder_pan_final_range_mean": 61.16154273351034, "shoulder_pan_final_range_median": 58.58602523803711, "pairwise_horizon_l2_mean": 30.735143926408554, "reference_horizon_l2_mean": 20.910792430241905, "pan_rank_match_rate": 1.0, "policy_config": { "type": "molmoact2", "chunk_size": 30, "image_keys": [ "observation.images.front" ], "norm_tag": "so100_so101_molmoact2", "action_mode": "continuous" }, "samples": [ { "index": 45085, "episode_index": 108, "frame_index": 205, "phase": 0.5352480417754569, "bin_index": 53, "target_position": "left", "target_color": "red", "prompt_bucket": "direct_color", "task": "Put the banana in the red bowl.", "source_repo": "rslxcvg/banana_blue", "source_episode": 18, "colors_by_position": "red,green,blue", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "red", "blue" ], "reference_decision_pan_rank": [ "red", "green", "blue" ], "reference_final_pan_rank": [ "red", "green", "blue" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "green": -2.4175825119018555, "blue": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "color_to_position": { "red": "left", "green": "center", "blue": "right" }, "reference_indices_by_prompt": { "red": 45085, "green": 91523, "blue": 144354 }, "pan_by_prompt": { "red": -5.538806915283203, "green": -0.14719390869140625, "blue": 4.819881439208984 }, "pan_final_by_prompt": { "red": -30.41663932800293, "green": -10.01766586303711, "blue": 25.707069396972656 }, "shoulder_pan_range": 10.358688354492188, "shoulder_pan_final_range": 56.123708724975586, "pairwise_horizon_l2_mean": 23.393242518107098, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 3.6594808101654053, "horizon_mean_l2": 19.036762237548828, "mean_abs_by_joint": { "shoulder_pan": 13.039280891418457, "shoulder_lift": 7.204833984375, "elbow_flex": 7.131406307220459, "wrist_flex": 8.336782455444336, "wrist_roll": 0.14615388214588165, "gripper": 0.097223661839962 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 3.1843068599700928, "horizon_mean_l2": 30.954692840576172, "mean_abs_by_joint": { "shoulder_pan": 30.32640266418457, "shoulder_lift": 2.315213918685913, "elbow_flex": 1.7054717540740967, "wrist_flex": 5.083708763122559, "wrist_roll": 0.30512815713882446, "gripper": 0.12805072963237762 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 2.144282579421997, "horizon_mean_l2": 20.18827247619629, "mean_abs_by_joint": { "shoulder_pan": 17.28712272644043, "shoulder_lift": 4.9156036376953125, "elbow_flex": 8.040634155273438, "wrist_flex": 3.2530741691589355, "wrist_roll": 0.25641027092933655, "gripper": 0.049797479063272476 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "red", "blue" ], "reference_decision_pan_rank": [ "red", "green", "blue" ], "reference_final_pan_rank": [ "red", "green", "blue" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "green": -2.4175825119018555, "blue": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "reference_pan_final_by_prompt": { "red": -5.626373767852783, "green": -2.1978023052215576, "blue": 19.604394912719727 }, "reference_final_pan_by_prompt": { "red": -31.120878219604492, "green": -6.593406677246094, "blue": 34.373626708984375 }, "predicted_final_pan_rank": [ "red", "green", "blue" ], "predicted_pan_rank": [ "red", "green", "blue" ], "reference_pan_rank": [ "red", "green", "blue" ], "reference_final_pan_rank": [ "red", "green", "blue" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 1.712888479232788, "reference_horizon_mean_l2": 1.4099116325378418, "reference_mean_abs_by_joint": { "shoulder_pan": 0.3438057005405426, "shoulder_lift": 0.8230550289154053, "elbow_flex": 0.739331841468811, "wrist_flex": 0.5064066648483276, "wrist_roll": 0.2604397237300873, "gripper": 0.0416795015335083 } }, "green": { "reference_first_action_l2": 32.64104461669922, "reference_horizon_mean_l2": 41.633338928222656, "reference_mean_abs_by_joint": { "shoulder_pan": 1.5724228620529175, "shoulder_lift": 25.479034423828125, "elbow_flex": 28.555387496948242, "wrist_flex": 11.94782829284668, "wrist_roll": 10.353846549987793, "gripper": 1.488412618637085 } }, "blue": { "reference_first_action_l2": 46.47682189941406, "reference_horizon_mean_l2": 42.82001876831055, "reference_mean_abs_by_joint": { "shoulder_pan": 12.935714721679688, "shoulder_lift": 12.053244590759277, "elbow_flex": 23.467418670654297, "wrist_flex": 29.131380081176758, "wrist_roll": 10.417582511901855, "gripper": 0.18310348689556122 } } }, "reference_horizon_l2_mean": 28.621089776357014 }, { "index": 45469, "episode_index": 109, "frame_index": 205, "phase": 0.5352480417754569, "bin_index": 53, "target_position": "left", "target_color": "red", "prompt_bucket": "direct_color", "task": "Put the banana in the red bowl.", "source_repo": "rslxcvg/banana_blue", "source_episode": 18, "colors_by_position": "red,blue,green", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "blue", "red", "green" ], "reference_decision_pan_rank": [ "red", "blue", "green" ], "reference_final_pan_rank": [ "red", "blue", "green" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "blue": -2.4175825119018555, "green": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "color_to_position": { "red": "left", "blue": "center", "green": "right" }, "reference_indices_by_prompt": { "red": 45469, "blue": 91949, "green": 144871 }, "pan_by_prompt": { "red": -5.560031890869141, "green": 5.414234161376953, "blue": -0.2745552062988281 }, "pan_final_by_prompt": { "red": -30.41663932800293, "green": 26.72595977783203, "blue": -8.510562896728516 }, "shoulder_pan_range": 10.974266052246094, "shoulder_pan_final_range": 57.14259910583496, "pairwise_horizon_l2_mean": 24.429942448933918, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 3.5566985607147217, "horizon_mean_l2": 31.749542236328125, "mean_abs_by_joint": { "shoulder_pan": 31.178659439086914, "shoulder_lift": 3.0947470664978027, "elbow_flex": 1.8211427927017212, "wrist_flex": 4.437337398529053, "wrist_roll": 0.26923075318336487, "gripper": 0.1209367886185646 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 4.117129802703857, "horizon_mean_l2": 20.2432861328125, "mean_abs_by_joint": { "shoulder_pan": 13.404027938842773, "shoulder_lift": 8.26077651977539, "elbow_flex": 7.6371283531188965, "wrist_flex": 8.861462593078613, "wrist_roll": 0.29487183690071106, "gripper": 0.06402534991502762 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 2.835785388946533, "horizon_mean_l2": 21.296998977661133, "mean_abs_by_joint": { "shoulder_pan": 17.774629592895508, "shoulder_lift": 5.176423072814941, "elbow_flex": 8.667407035827637, "wrist_flex": 4.424124240875244, "wrist_roll": 0.29230767488479614, "gripper": 0.06165403127670288 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "blue", "red", "green" ], "reference_decision_pan_rank": [ "red", "blue", "green" ], "reference_final_pan_rank": [ "red", "blue", "green" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "blue": -2.4175825119018555, "green": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "reference_pan_final_by_prompt": { "red": -5.626373767852783, "blue": -2.1978023052215576, "green": 19.604394912719727 }, "reference_final_pan_by_prompt": { "red": -31.120878219604492, "blue": -6.593406677246094, "green": 34.373626708984375 }, "predicted_final_pan_rank": [ "red", "blue", "green" ], "predicted_pan_rank": [ "red", "blue", "green" ], "reference_pan_rank": [ "red", "blue", "green" ], "reference_final_pan_rank": [ "red", "blue", "green" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 1.5516172647476196, "reference_horizon_mean_l2": 1.0013529062271118, "reference_mean_abs_by_joint": { "shoulder_pan": 0.5059507489204407, "shoulder_lift": 0.5388623476028442, "elbow_flex": 0.3704335391521454, "wrist_flex": 0.22744546830654144, "wrist_roll": 0.2007327526807785, "gripper": 0.04375172033905983 } }, "green": { "reference_first_action_l2": 45.9112434387207, "reference_horizon_mean_l2": 41.05442810058594, "reference_mean_abs_by_joint": { "shoulder_pan": 11.960697174072266, "shoulder_lift": 10.772212982177734, "elbow_flex": 22.95093536376953, "wrist_flex": 27.949207305908203, "wrist_roll": 10.481685638427734, "gripper": 0.1878460943698883 } }, "blue": { "reference_first_action_l2": 32.86920928955078, "reference_horizon_mean_l2": 39.93840789794922, "reference_mean_abs_by_joint": { "shoulder_pan": 1.2745020389556885, "shoulder_lift": 23.93718147277832, "elbow_flex": 27.57352638244629, "wrist_flex": 11.936701774597168, "wrist_roll": 10.24871826171875, "gripper": 1.5097543001174927 } } }, "reference_horizon_l2_mean": 27.33139630158742 }, { "index": 46620, "episode_index": 112, "frame_index": 204, "phase": 0.5326370757180157, "bin_index": 53, "target_position": "left", "target_color": "blue", "prompt_bucket": "direct_color", "task": "Put the banana in the blue bowl.", "source_repo": "rslxcvg/banana_blue", "source_episode": 18, "colors_by_position": "blue,red,green", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 14.94505524635315, "reference_decision_pan_range": 23.03296661376953, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "red", "blue", "green" ], "reference_decision_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "reference_first_l2_mean": 34.4804433186849, "reference_early_l2_mean": 37.67904917399088, "reference_horizon_pair_l2_mean": 49.746201833089195, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "red": -2.4175825119018555, "green": 8.659340858459473 }, "reference_state_pan_range": 11.076923370361328, "reference_state_l2_mean": 33.57865333557129 }, "color_to_position": { "blue": "left", "red": "center", "green": "right" }, "reference_indices_by_prompt": { "blue": 46620, "red": 93226, "green": 146421 }, "pan_by_prompt": { "red": 0.319793701171875, "green": 3.9708099365234375, "blue": -3.925567626953125 }, "pan_final_by_prompt": { "red": -9.741718292236328, "green": 25.707069396972656, "blue": -31.010990142822266 }, "shoulder_pan_range": 7.8963775634765625, "shoulder_pan_final_range": 56.71805953979492, "pairwise_horizon_l2_mean": 23.31292215983073, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 3.2059004306793213, "horizon_mean_l2": 20.576030731201172, "mean_abs_by_joint": { "shoulder_pan": 16.50809669494629, "shoulder_lift": 4.521940231323242, "elbow_flex": 7.68941593170166, "wrist_flex": 6.294745922088623, "wrist_roll": 0.6846153140068054, "gripper": 0.08536709100008011 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 3.7863733768463135, "horizon_mean_l2": 19.684839248657227, "mean_abs_by_joint": { "shoulder_pan": 12.716278076171875, "shoulder_lift": 7.849247455596924, "elbow_flex": 6.63493013381958, "wrist_flex": 9.741313934326172, "wrist_roll": 0.5846153497695923, "gripper": 0.07825320214033127 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 2.1443941593170166, "horizon_mean_l2": 29.67789649963379, "mean_abs_by_joint": { "shoulder_pan": 29.224376678466797, "shoulder_lift": 3.431246280670166, "elbow_flex": 1.285827875137329, "wrist_flex": 3.4632568359375, "wrist_roll": 0.2999999225139618, "gripper": 0.16362029314041138 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 14.94505524635315, "reference_decision_pan_range": 23.03296661376953, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "red", "blue", "green" ], "reference_decision_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "reference_first_l2_mean": 34.4804433186849, "reference_early_l2_mean": 37.67904917399088, "reference_horizon_pair_l2_mean": 49.746201833089195, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "red": -2.4175825119018555, "green": 8.659340858459473 }, "reference_state_pan_range": 11.076923370361328, "reference_state_l2_mean": 33.57865333557129 }, "reference_pan_final_by_prompt": { "blue": -4.483516693115234, "red": -2.1978023052215576, "green": 18.549449920654297 }, "reference_final_pan_by_prompt": { "blue": -30.41758155822754, "red": -6.593406677246094, "green": 34.373626708984375 }, "predicted_final_pan_rank": [ "blue", "red", "green" ], "predicted_pan_rank": [ "blue", "red", "green" ], "reference_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 31.029787063598633, "reference_horizon_mean_l2": 38.99106216430664, "reference_mean_abs_by_joint": { "shoulder_pan": 1.6803377866744995, "shoulder_lift": 23.847566604614258, "elbow_flex": 26.37550926208496, "wrist_flex": 11.867459297180176, "wrist_roll": 9.717949867248535, "gripper": 1.4619650840759277 } }, "green": { "reference_first_action_l2": 44.49510192871094, "reference_horizon_mean_l2": 39.1322021484375, "reference_mean_abs_by_joint": { "shoulder_pan": 12.797966957092285, "shoulder_lift": 10.630517959594727, "elbow_flex": 21.280675888061523, "wrist_flex": 26.01226806640625, "wrist_roll": 10.634066581726074, "gripper": 0.2613564431667328 } }, "blue": { "reference_first_action_l2": 2.754443407058716, "reference_horizon_mean_l2": 2.628492593765259, "reference_mean_abs_by_joint": { "shoulder_pan": 0.357387512922287, "shoulder_lift": 1.0190844535827637, "elbow_flex": 1.704785704612732, "wrist_flex": 1.396203637123108, "wrist_roll": 0.20366282761096954, "gripper": 0.03894500806927681 } } }, "reference_horizon_l2_mean": 26.9172523021698 }, { "index": 47005, "episode_index": 113, "frame_index": 205, "phase": 0.5352480417754569, "bin_index": 53, "target_position": "left", "target_color": "blue", "prompt_bucket": "direct_color", "task": "Put the banana in the blue bowl.", "source_repo": "rslxcvg/banana_blue", "source_episode": 18, "colors_by_position": "blue,green,red", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "blue", "red" ], "reference_decision_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "green": -2.4175825119018555, "red": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "color_to_position": { "blue": "left", "green": "center", "red": "right" }, "reference_indices_by_prompt": { "blue": 47005, "green": 93653, "red": 146939 }, "pan_by_prompt": { "red": 4.650066375732422, "green": -0.019832611083984375, "blue": -4.562374114990234 }, "pan_final_by_prompt": { "red": 26.55614471435547, "green": -9.444541931152344, "blue": -30.33173179626465 }, "shoulder_pan_range": 9.212440490722656, "shoulder_pan_final_range": 56.88787651062012, "pairwise_horizon_l2_mean": 23.258761723836262, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 2.4635329246520996, "horizon_mean_l2": 20.048696517944336, "mean_abs_by_joint": { "shoulder_pan": 16.750261306762695, "shoulder_lift": 4.623605728149414, "elbow_flex": 6.799860954284668, "wrist_flex": 5.968605041503906, "wrist_roll": 0.4205128252506256, "gripper": 0.035569626837968826 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 2.996840238571167, "horizon_mean_l2": 30.273733139038086, "mean_abs_by_joint": { "shoulder_pan": 29.85835075378418, "shoulder_lift": 2.941439390182495, "elbow_flex": 1.1378788948059082, "wrist_flex": 3.420315980911255, "wrist_roll": 0.5999998450279236, "gripper": 0.12093671411275864 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 3.9923877716064453, "horizon_mean_l2": 19.453855514526367, "mean_abs_by_joint": { "shoulder_pan": 13.108091354370117, "shoulder_lift": 7.513075351715088, "elbow_flex": 6.3344879150390625, "wrist_flex": 9.388920783996582, "wrist_roll": 0.3641025125980377, "gripper": 0.11382277309894562 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "blue", "red" ], "reference_decision_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "green": -2.4175825119018555, "red": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "reference_pan_final_by_prompt": { "blue": -5.626373767852783, "green": -2.1978023052215576, "red": 19.604394912719727 }, "reference_final_pan_by_prompt": { "blue": -31.120878219604492, "green": -6.593406677246094, "red": 34.373626708984375 }, "predicted_final_pan_rank": [ "blue", "green", "red" ], "predicted_pan_rank": [ "blue", "green", "red" ], "reference_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 46.04142379760742, "reference_horizon_mean_l2": 39.85016632080078, "reference_mean_abs_by_joint": { "shoulder_pan": 12.81259822845459, "shoulder_lift": 12.643092155456543, "elbow_flex": 21.404176712036133, "wrist_flex": 26.154382705688477, "wrist_roll": 10.271428108215332, "gripper": 0.055052749812603 } }, "green": { "reference_first_action_l2": 32.678138732910156, "reference_horizon_mean_l2": 42.052547454833984, "reference_mean_abs_by_joint": { "shoulder_pan": 1.72751784324646, "shoulder_lift": 26.376466751098633, "elbow_flex": 28.507631301879883, "wrist_flex": 11.686358451843262, "wrist_roll": 10.120513916015625, "gripper": 1.5927501916885376 } }, "blue": { "reference_first_action_l2": 1.8410029411315918, "reference_horizon_mean_l2": 2.6852669715881348, "reference_mean_abs_by_joint": { "shoulder_pan": 0.9041402339935303, "shoulder_lift": 1.7069882154464722, "elbow_flex": 1.2076936960220337, "wrist_flex": 0.8824427127838135, "wrist_roll": 0.27069610357284546, "gripper": 0.14791837334632874 } } }, "reference_horizon_l2_mean": 28.195993582407635 }, { "index": 91524, "episode_index": 228, "frame_index": 228, "phase": 0.5364705882352941, "bin_index": 53, "target_position": "center", "target_color": "green", "prompt_bucket": "direct_color", "task": "Put the banana in the green bowl.", "source_repo": "rslxcvg/banana_red1", "source_episode": 18, "colors_by_position": "red,green,blue", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 17.230769395828247, "reference_decision_pan_range": 26.28571367263794, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "red", "blue" ], "reference_decision_pan_rank": [ "red", "green", "blue" ], "reference_final_pan_rank": [ "red", "green", "blue" ], "reference_first_l2_mean": 35.64093271891276, "reference_early_l2_mean": 39.67741139729818, "reference_horizon_pair_l2_mean": 51.860599517822266, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "green": -2.4175825119018555, "blue": 10.94505500793457 }, "reference_state_pan_range": 13.362637519836426, "reference_state_l2_mean": 34.12602170308431 }, "color_to_position": { "red": "left", "green": "center", "blue": "right" }, "reference_indices_by_prompt": { "red": 45085, "green": 91524, "blue": 144355 }, "pan_by_prompt": { "red": -9.688648223876953, "green": -1.8877944946289062, "blue": 8.046356201171875 }, "pan_final_by_prompt": { "red": -33.21858215332031, "green": -6.642601013183594, "blue": 25.027816772460938 }, "shoulder_pan_range": 17.735004425048828, "shoulder_pan_final_range": 58.24639892578125, "pairwise_horizon_l2_mean": 30.988487243652344, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 4.828479766845703, "horizon_mean_l2": 29.193815231323242, "mean_abs_by_joint": { "shoulder_pan": 16.02837371826172, "shoulder_lift": 19.426774978637695, "elbow_flex": 12.62291145324707, "wrist_flex": 6.841937065124512, "wrist_roll": 0.9371795654296875, "gripper": 0.5951985716819763 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 7.1735968589782715, "horizon_mean_l2": 35.65236282348633, "mean_abs_by_joint": { "shoulder_pan": 35.18274688720703, "shoulder_lift": 4.816863536834717, "elbow_flex": 0.5447283387184143, "wrist_flex": 2.453364849090576, "wrist_roll": 1.2141029834747314, "gripper": 0.5928269624710083 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 5.541849136352539, "horizon_mean_l2": 28.11928367614746, "mean_abs_by_joint": { "shoulder_pan": 19.154375076293945, "shoulder_lift": 14.60991096496582, "elbow_flex": 13.167641639709473, "wrist_flex": 4.410825729370117, "wrist_roll": 2.151282787322998, "gripper": 1.1880255937576294 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 17.230769395828247, "reference_decision_pan_range": 26.28571367263794, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "red", "blue" ], "reference_decision_pan_rank": [ "red", "green", "blue" ], "reference_final_pan_rank": [ "red", "green", "blue" ], "reference_first_l2_mean": 35.64093271891276, "reference_early_l2_mean": 39.67741139729818, "reference_horizon_pair_l2_mean": 51.860599517822266, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "green": -2.4175825119018555, "blue": 10.94505500793457 }, "reference_state_pan_range": 13.362637519836426, "reference_state_l2_mean": 34.12602170308431 }, "reference_pan_final_by_prompt": { "red": -5.626373767852783, "green": -2.1978023052215576, "blue": 20.659339904785156 }, "reference_final_pan_by_prompt": { "red": -31.120878219604492, "green": -6.593406677246094, "blue": 34.373626708984375 }, "predicted_final_pan_rank": [ "red", "green", "blue" ], "predicted_pan_rank": [ "red", "green", "blue" ], "reference_pan_rank": [ "red", "green", "blue" ], "reference_final_pan_rank": [ "red", "green", "blue" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 29.991186141967773, "reference_horizon_mean_l2": 23.848310470581055, "reference_mean_abs_by_joint": { "shoulder_pan": 3.0051329135894775, "shoulder_lift": 8.577493667602539, "elbow_flex": 17.119150161743164, "wrist_flex": 9.742671966552734, "wrist_roll": 9.759798049926758, "gripper": 1.1705288887023926 } }, "green": { "reference_first_action_l2": 2.2710914611816406, "reference_horizon_mean_l2": 7.577295303344727, "reference_mean_abs_by_joint": { "shoulder_pan": 0.2420756220817566, "shoulder_lift": 5.396466255187988, "elbow_flex": 5.264532089233398, "wrist_flex": 0.27254435420036316, "wrist_roll": 0.20320479571819305, "gripper": 0.2070728987455368 } }, "blue": { "reference_first_action_l2": 23.25484848022461, "reference_horizon_mean_l2": 21.11150360107422, "reference_mean_abs_by_joint": { "shoulder_pan": 12.074399948120117, "shoulder_lift": 1.5728487968444824, "elbow_flex": 4.896638870239258, "wrist_flex": 16.29283905029297, "wrist_roll": 1.8777481317520142, "gripper": 0.6597363352775574 } } }, "reference_horizon_l2_mean": 17.512369791666668 }, { "index": 91950, "episode_index": 229, "frame_index": 228, "phase": 0.5364705882352941, "bin_index": 53, "target_position": "center", "target_color": "blue", "prompt_bucket": "direct_color", "task": "Put the banana in the blue bowl.", "source_repo": "rslxcvg/banana_red1", "source_episode": 18, "colors_by_position": "red,blue,green", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 17.230769395828247, "reference_decision_pan_range": 26.28571367263794, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "blue", "red", "green" ], "reference_decision_pan_rank": [ "red", "blue", "green" ], "reference_final_pan_rank": [ "red", "blue", "green" ], "reference_first_l2_mean": 35.64093271891276, "reference_early_l2_mean": 39.67741139729818, "reference_horizon_pair_l2_mean": 51.860599517822266, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "blue": -2.4175825119018555, "green": 10.94505500793457 }, "reference_state_pan_range": 13.362637519836426, "reference_state_l2_mean": 34.12602170308431 }, "color_to_position": { "red": "left", "blue": "center", "green": "right" }, "reference_indices_by_prompt": { "red": 45469, "blue": 91950, "green": 144872 }, "pan_by_prompt": { "red": -11.397407531738281, "green": 8.470893859863281, "blue": -2.312328338623047 }, "pan_final_by_prompt": { "red": -34.40727996826172, "green": 24.51837158203125, "blue": -7.130817413330078 }, "shoulder_pan_range": 19.868301391601562, "shoulder_pan_final_range": 58.92565155029297, "pairwise_horizon_l2_mean": 34.69123967488607, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 8.831618309020996, "horizon_mean_l2": 37.170406341552734, "mean_abs_by_joint": { "shoulder_pan": 36.48430633544922, "shoulder_lift": 5.1275410652160645, "elbow_flex": 0.5877711176872253, "wrist_flex": 2.6807596683502197, "wrist_roll": 3.196153402328491, "gripper": 0.11619415134191513 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 7.813859939575195, "horizon_mean_l2": 33.897342681884766, "mean_abs_by_joint": { "shoulder_pan": 16.576021194458008, "shoulder_lift": 23.322978973388672, "elbow_flex": 16.23963737487793, "wrist_flex": 6.725891590118408, "wrist_roll": 4.200000286102295, "gripper": 1.0338904857635498 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 8.816144943237305, "horizon_mean_l2": 33.0059700012207, "mean_abs_by_joint": { "shoulder_pan": 19.908281326293945, "shoulder_lift": 18.195436477661133, "elbow_flex": 16.827411651611328, "wrist_flex": 4.045132637023926, "wrist_roll": 7.396154403686523, "gripper": 1.1453418731689453 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 17.230769395828247, "reference_decision_pan_range": 26.28571367263794, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "blue", "red", "green" ], "reference_decision_pan_rank": [ "red", "blue", "green" ], "reference_final_pan_rank": [ "red", "blue", "green" ], "reference_first_l2_mean": 35.64093271891276, "reference_early_l2_mean": 39.67741139729818, "reference_horizon_pair_l2_mean": 51.860599517822266, "reference_state_pan_by_prompt": { "red": -0.3956044018268585, "blue": -2.4175825119018555, "green": 10.94505500793457 }, "reference_state_pan_range": 13.362637519836426, "reference_state_l2_mean": 34.12602170308431 }, "reference_pan_final_by_prompt": { "red": -5.626373767852783, "blue": -2.1978023052215576, "green": 20.659339904785156 }, "reference_final_pan_by_prompt": { "red": -31.120878219604492, "blue": -6.593406677246094, "green": 34.373626708984375 }, "predicted_final_pan_rank": [ "red", "blue", "green" ], "predicted_pan_rank": [ "red", "blue", "green" ], "reference_pan_rank": [ "red", "blue", "green" ], "reference_final_pan_rank": [ "red", "blue", "green" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 30.24158477783203, "reference_horizon_mean_l2": 24.185546875, "reference_mean_abs_by_joint": { "shoulder_pan": 4.0208353996276855, "shoulder_lift": 10.619543075561523, "elbow_flex": 17.73785400390625, "wrist_flex": 10.067767143249512, "wrist_roll": 6.0565924644470215, "gripper": 0.2694319784641266 } }, "green": { "reference_first_action_l2": 23.883623123168945, "reference_horizon_mean_l2": 21.98788070678711, "reference_mean_abs_by_joint": { "shoulder_pan": 11.788546562194824, "shoulder_lift": 1.9405322074890137, "elbow_flex": 4.3209757804870605, "wrist_flex": 16.217388153076172, "wrist_roll": 7.563004016876221, "gripper": 0.24001480638980865 } }, "blue": { "reference_first_action_l2": 2.417416572570801, "reference_horizon_mean_l2": 1.6627991199493408, "reference_mean_abs_by_joint": { "shoulder_pan": 0.3201746642589569, "shoulder_lift": 0.6046574711799622, "elbow_flex": 1.2546254396438599, "wrist_flex": 0.22125473618507385, "wrist_roll": 0.23717989027500153, "gripper": 0.2553320825099945 } } }, "reference_horizon_l2_mean": 15.945408900578817 }, { "index": 93226, "episode_index": 232, "frame_index": 226, "phase": 0.5317647058823529, "bin_index": 53, "target_position": "center", "target_color": "red", "prompt_bucket": "direct_color", "task": "Put the banana in the red bowl.", "source_repo": "rslxcvg/banana_red1", "source_episode": 18, "colors_by_position": "blue,red,green", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 13.802197694778442, "reference_decision_pan_range": 21.89011001586914, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "red", "blue", "green" ], "reference_decision_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "reference_first_l2_mean": 34.164158503214516, "reference_early_l2_mean": 37.280344009399414, "reference_horizon_pair_l2_mean": 49.451089223225914, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "red": -2.4175825119018555, "green": 7.516483306884766 }, "reference_state_pan_range": 9.934065818786621, "reference_state_l2_mean": 33.34150505065918 }, "color_to_position": { "blue": "left", "red": "center", "green": "right" }, "reference_indices_by_prompt": { "blue": 46620, "red": 93226, "green": 146420 }, "pan_by_prompt": { "red": -2.2274246215820312, "green": 8.046356201171875, "blue": -11.100234985351562 }, "pan_final_by_prompt": { "red": -6.345424652099609, "green": 24.858001708984375, "blue": -34.067649841308594 }, "shoulder_pan_range": 19.146591186523438, "shoulder_pan_final_range": 58.92565155029297, "pairwise_horizon_l2_mean": 33.44681866963705, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 6.295548915863037, "horizon_mean_l2": 31.358232498168945, "mean_abs_by_joint": { "shoulder_pan": 19.490467071533203, "shoulder_lift": 17.704492568969727, "elbow_flex": 16.366071701049805, "wrist_flex": 2.8313136100769043, "wrist_roll": 2.6660258769989014, "gripper": 0.9532660841941833 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 5.750476360321045, "horizon_mean_l2": 32.13639450073242, "mean_abs_by_joint": { "shoulder_pan": 16.864709854125977, "shoulder_lift": 21.55246353149414, "elbow_flex": 15.45953369140625, "wrist_flex": 6.291615962982178, "wrist_roll": 0.5743591785430908, "gripper": 0.26084384322166443 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 8.136035919189453, "horizon_mean_l2": 36.845829010009766, "mean_abs_by_joint": { "shoulder_pan": 36.35517120361328, "shoulder_lift": 3.8479695320129395, "elbow_flex": 0.9065371155738831, "wrist_flex": 3.4603030681610107, "wrist_roll": 2.0916666984558105, "gripper": 1.1998820304870605 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 13.802197694778442, "reference_decision_pan_range": 21.89011001586914, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "red", "blue", "green" ], "reference_decision_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "reference_first_l2_mean": 34.164158503214516, "reference_early_l2_mean": 37.280344009399414, "reference_horizon_pair_l2_mean": 49.451089223225914, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "red": -2.4175825119018555, "green": 7.516483306884766 }, "reference_state_pan_range": 9.934065818786621, "reference_state_l2_mean": 33.34150505065918 }, "reference_pan_final_by_prompt": { "blue": -4.483516693115234, "red": -2.1978023052215576, "green": 17.406593322753906 }, "reference_final_pan_by_prompt": { "blue": -30.41758155822754, "red": -6.593406677246094, "green": 34.373626708984375 }, "predicted_final_pan_rank": [ "blue", "red", "green" ], "predicted_pan_rank": [ "blue", "red", "green" ], "reference_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 1.9544914960861206, "reference_horizon_mean_l2": 1.3550655841827393, "reference_mean_abs_by_joint": { "shoulder_pan": 0.23658305406570435, "shoulder_lift": 0.8437034487724304, "elbow_flex": 0.6412086486816406, "wrist_flex": 0.4646644592285156, "wrist_roll": 0.15705108642578125, "gripper": 0.20517149567604065 } }, "green": { "reference_first_action_l2": 22.214157104492188, "reference_horizon_mean_l2": 21.257524490356445, "reference_mean_abs_by_joint": { "shoulder_pan": 9.623479843139648, "shoulder_lift": 1.2871012687683105, "elbow_flex": 4.32573127746582, "wrist_flex": 18.07876968383789, "wrist_roll": 2.9082424640655518, "gripper": 0.47951698303222656 } }, "blue": { "reference_first_action_l2": 30.776382446289062, "reference_horizon_mean_l2": 25.70706558227539, "reference_mean_abs_by_joint": { "shoulder_pan": 4.89147424697876, "shoulder_lift": 10.455570220947266, "elbow_flex": 18.283367156982422, "wrist_flex": 9.349300384521484, "wrist_roll": 10.058148384094238, "gripper": 1.5973644256591797 } } }, "reference_horizon_l2_mean": 16.10655188560486 }, { "index": 93654, "episode_index": 233, "frame_index": 228, "phase": 0.5364705882352941, "bin_index": 53, "target_position": "center", "target_color": "green", "prompt_bucket": "direct_color", "task": "Put the banana in the green bowl.", "source_repo": "rslxcvg/banana_red1", "source_episode": 18, "colors_by_position": "blue,green,red", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 17.230769395828247, "reference_decision_pan_range": 26.28571367263794, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "blue", "red" ], "reference_decision_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "reference_first_l2_mean": 35.64093271891276, "reference_early_l2_mean": 39.67741139729818, "reference_horizon_pair_l2_mean": 51.860599517822266, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "green": -2.4175825119018555, "red": 10.94505500793457 }, "reference_state_pan_range": 13.362637519836426, "reference_state_l2_mean": 34.12602170308431 }, "color_to_position": { "blue": "left", "green": "center", "red": "right" }, "reference_indices_by_prompt": { "blue": 47005, "green": 93654, "red": 146940 }, "pan_by_prompt": { "red": 7.876544952392578, "green": -1.802886962890625, "blue": -10.038890838623047 }, "pan_final_by_prompt": { "red": 25.707069396972656, "green": -6.38787841796875, "blue": -32.029876708984375 }, "shoulder_pan_range": 17.915435791015625, "shoulder_pan_final_range": 57.73694610595703, "pairwise_horizon_l2_mean": 32.68465105692545, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 5.743428707122803, "horizon_mean_l2": 31.264516830444336, "mean_abs_by_joint": { "shoulder_pan": 19.31852912902832, "shoulder_lift": 17.62553596496582, "elbow_flex": 16.035198211669922, "wrist_flex": 3.7900519371032715, "wrist_roll": 3.9006412029266357, "gripper": 1.2188526391983032 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 7.439505100250244, "horizon_mean_l2": 35.34449768066406, "mean_abs_by_joint": { "shoulder_pan": 34.90927505493164, "shoulder_lift": 3.825913667678833, "elbow_flex": 0.9186411499977112, "wrist_flex": 3.0427145957946777, "wrist_roll": 1.8724359273910522, "gripper": 0.6900507211685181 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 5.219200611114502, "horizon_mean_l2": 31.44493865966797, "mean_abs_by_joint": { "shoulder_pan": 15.590744972229004, "shoulder_lift": 21.451452255249023, "elbow_flex": 15.116559028625488, "wrist_flex": 6.832767009735107, "wrist_roll": 2.028205394744873, "gripper": 0.5288017988204956 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 17.230769395828247, "reference_decision_pan_range": 26.28571367263794, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "blue", "red" ], "reference_decision_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "reference_first_l2_mean": 35.64093271891276, "reference_early_l2_mean": 39.67741139729818, "reference_horizon_pair_l2_mean": 51.860599517822266, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "green": -2.4175825119018555, "red": 10.94505500793457 }, "reference_state_pan_range": 13.362637519836426, "reference_state_l2_mean": 34.12602170308431 }, "reference_pan_final_by_prompt": { "blue": -5.626373767852783, "green": -2.1978023052215576, "red": 20.659339904785156 }, "reference_final_pan_by_prompt": { "blue": -31.120878219604492, "green": -6.593406677246094, "red": 34.373626708984375 }, "predicted_final_pan_rank": [ "blue", "green", "red" ], "predicted_pan_rank": [ "blue", "green", "red" ], "reference_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 23.669963836669922, "reference_horizon_mean_l2": 21.202260971069336, "reference_mean_abs_by_joint": { "shoulder_pan": 11.675337791442871, "shoulder_lift": 1.8304036855697632, "elbow_flex": 4.409745693206787, "wrist_flex": 16.501806259155273, "wrist_roll": 3.8123629093170166, "gripper": 0.33960965275764465 } }, "green": { "reference_first_action_l2": 1.3375098705291748, "reference_horizon_mean_l2": 3.431023359298706, "reference_mean_abs_by_joint": { "shoulder_pan": 0.42244675755500793, "shoulder_lift": 2.729001045227051, "elbow_flex": 1.9100794792175293, "wrist_flex": 0.19077707827091217, "wrist_roll": 0.06410261988639832, "gripper": 0.09376271069049835 } }, "blue": { "reference_first_action_l2": 30.042041778564453, "reference_horizon_mean_l2": 24.351747512817383, "reference_mean_abs_by_joint": { "shoulder_pan": 2.3325958251953125, "shoulder_lift": 9.220279693603516, "elbow_flex": 17.979955673217773, "wrist_flex": 10.145306587219238, "wrist_roll": 8.483515739440918, "gripper": 0.9476262927055359 } } }, "reference_horizon_l2_mean": 16.328343947728474 }, { "index": 145388, "episode_index": 350, "frame_index": 276, "phase": 0.5348837209302325, "bin_index": 53, "target_position": "right", "target_color": "blue", "prompt_bucket": "direct_color", "task": "Put the banana in the blue bowl.", "source_repo": "rslxcvg/banana_green", "source_episode": 18, "colors_by_position": "green,red,blue", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "red", "green", "blue" ], "reference_decision_pan_rank": [ "green", "red", "blue" ], "reference_final_pan_rank": [ "green", "red", "blue" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "green": -0.3956044018268585, "red": -2.4175825119018555, "blue": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "color_to_position": { "green": "left", "red": "center", "blue": "right" }, "reference_indices_by_prompt": { "green": 45853, "red": 92375, "blue": 145388 }, "pan_by_prompt": { "red": 2.612293243408203, "green": -11.07900619506836, "blue": 19.423934936523438 }, "pan_final_by_prompt": { "red": -6.515239715576172, "green": -34.746910095214844, "blue": 34.19779968261719 }, "shoulder_pan_range": 30.502941131591797, "shoulder_pan_final_range": 68.94470977783203, "pairwise_horizon_l2_mean": 36.17309697469076, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 7.254165172576904, "horizon_mean_l2": 26.775718688964844, "mean_abs_by_joint": { "shoulder_pan": 20.263118743896484, "shoulder_lift": 9.020495414733887, "elbow_flex": 11.347545623779297, "wrist_flex": 4.924811840057373, "wrist_roll": 5.802885055541992, "gripper": 0.1991899013519287 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 6.799628257751465, "horizon_mean_l2": 31.572471618652344, "mean_abs_by_joint": { "shoulder_pan": 29.232515335083008, "shoulder_lift": 6.399641036987305, "elbow_flex": 6.955040454864502, "wrist_flex": 5.862208843231201, "wrist_roll": 0.9124199748039246, "gripper": 0.06876792758703232 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 13.278752326965332, "horizon_mean_l2": 50.17110061645508, "mean_abs_by_joint": { "shoulder_pan": 49.49563217163086, "shoulder_lift": 2.620854139328003, "elbow_flex": 4.443615913391113, "wrist_flex": 0.937396764755249, "wrist_roll": 5.409696102142334, "gripper": 0.17310555279254913 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "red", "green", "blue" ], "reference_decision_pan_rank": [ "green", "red", "blue" ], "reference_final_pan_rank": [ "green", "red", "blue" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "green": -0.3956044018268585, "red": -2.4175825119018555, "blue": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "reference_pan_final_by_prompt": { "green": -5.626373767852783, "red": -2.1978023052215576, "blue": 19.604394912719727 }, "reference_final_pan_by_prompt": { "green": -31.120878219604492, "red": -6.593406677246094, "blue": 34.373626708984375 }, "predicted_final_pan_rank": [ "green", "red", "blue" ], "predicted_pan_rank": [ "green", "red", "blue" ], "reference_pan_rank": [ "green", "red", "blue" ], "reference_final_pan_rank": [ "green", "red", "blue" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 22.611278533935547, "reference_horizon_mean_l2": 20.9271297454834, "reference_mean_abs_by_joint": { "shoulder_pan": 2.3782715797424316, "shoulder_lift": 11.683895111083984, "elbow_flex": 6.461350440979004, "wrist_flex": 15.051709175109863, "wrist_roll": 1.6128199100494385, "gripper": 1.3319060802459717 } }, "green": { "reference_first_action_l2": 43.3091926574707, "reference_horizon_mean_l2": 32.57255172729492, "reference_mean_abs_by_joint": { "shoulder_pan": 6.188817501068115, "shoulder_lift": 11.63553524017334, "elbow_flex": 16.78857421875, "wrist_flex": 23.153079986572266, "wrist_roll": 6.303708553314209, "gripper": 0.1014321893453598 } }, "blue": { "reference_first_action_l2": 0.7755581736564636, "reference_horizon_mean_l2": 1.6767138242721558, "reference_mean_abs_by_joint": { "shoulder_pan": 0.23671665787696838, "shoulder_lift": 0.5040987730026245, "elbow_flex": 0.35540109872817993, "wrist_flex": 0.4772895872592926, "wrist_roll": 1.2479965686798096, "gripper": 0.24950002133846283 } } }, "reference_horizon_l2_mean": 18.39213176568349 }, { "index": 145904, "episode_index": 351, "frame_index": 275, "phase": 0.5329457364341085, "bin_index": 53, "target_position": "right", "target_color": "red", "prompt_bucket": "direct_color", "task": "Put the banana in the red bowl.", "source_repo": "rslxcvg/banana_green", "source_episode": 18, "colors_by_position": "green,blue,red", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 14.94505524635315, "reference_decision_pan_range": 23.03296661376953, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "blue", "green", "red" ], "reference_decision_pan_rank": [ "green", "blue", "red" ], "reference_final_pan_rank": [ "green", "blue", "red" ], "reference_first_l2_mean": 34.76588694254557, "reference_early_l2_mean": 38.33881123860677, "reference_horizon_pair_l2_mean": 50.437808990478516, "reference_state_pan_by_prompt": { "green": -0.3956044018268585, "blue": -2.4175825119018555, "red": 8.659340858459473 }, "reference_state_pan_range": 11.076923370361328, "reference_state_l2_mean": 33.57865333557129 }, "color_to_position": { "green": "left", "blue": "center", "red": "right" }, "reference_indices_by_prompt": { "green": 46236, "blue": 92801, "red": 145904 }, "pan_by_prompt": { "red": 18.405044555664062, "green": -9.33840560913086, "blue": 1.33868408203125 }, "pan_final_by_prompt": { "red": 33.51853942871094, "green": -34.40727996826172, "blue": -5.793529510498047 }, "shoulder_pan_range": 27.743450164794922, "shoulder_pan_final_range": 67.92581939697266, "pairwise_horizon_l2_mean": 35.69016202290853, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 11.474370002746582, "horizon_mean_l2": 48.006813049316406, "mean_abs_by_joint": { "shoulder_pan": 47.50455856323242, "shoulder_lift": 2.5555686950683594, "elbow_flex": 3.579572916030884, "wrist_flex": 0.5535356998443604, "wrist_roll": 4.402884483337402, "gripper": 0.19918985664844513 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 6.814485549926758, "horizon_mean_l2": 32.852317810058594, "mean_abs_by_joint": { "shoulder_pan": 29.41259002685547, "shoulder_lift": 8.466378211975098, "elbow_flex": 10.294698715209961, "wrist_flex": 3.7384603023529053, "wrist_roll": 1.3104166984558105, "gripper": 0.15176373720169067 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 5.366869926452637, "horizon_mean_l2": 26.211355209350586, "mean_abs_by_joint": { "shoulder_pan": 18.09197235107422, "shoulder_lift": 11.021944999694824, "elbow_flex": 13.807022094726562, "wrist_flex": 3.184924602508545, "wrist_roll": 3.6985573768615723, "gripper": 0.06165401265025139 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 14.94505524635315, "reference_decision_pan_range": 23.03296661376953, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "blue", "green", "red" ], "reference_decision_pan_rank": [ "green", "blue", "red" ], "reference_final_pan_rank": [ "green", "blue", "red" ], "reference_first_l2_mean": 34.76588694254557, "reference_early_l2_mean": 38.33881123860677, "reference_horizon_pair_l2_mean": 50.437808990478516, "reference_state_pan_by_prompt": { "green": -0.3956044018268585, "blue": -2.4175825119018555, "red": 8.659340858459473 }, "reference_state_pan_range": 11.076923370361328, "reference_state_l2_mean": 33.57865333557129 }, "reference_pan_final_by_prompt": { "green": -4.483516693115234, "blue": -2.1978023052215576, "red": 18.549449920654297 }, "reference_final_pan_by_prompt": { "green": -30.41758155822754, "blue": -6.593406677246094, "red": 34.373626708984375 }, "predicted_final_pan_rank": [ "green", "blue", "red" ], "predicted_pan_rank": [ "green", "blue", "red" ], "reference_pan_rank": [ "green", "blue", "red" ], "reference_final_pan_rank": [ "green", "blue", "red" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 0.9038380980491638, "reference_horizon_mean_l2": 1.347490906715393, "reference_mean_abs_by_joint": { "shoulder_pan": 0.4050765335559845, "shoulder_lift": 0.617021918296814, "elbow_flex": 0.4057164490222931, "wrist_flex": 0.6882418394088745, "wrist_roll": 0.47000905871391296, "gripper": 0.03826131299138069 } }, "green": { "reference_first_action_l2": 44.467254638671875, "reference_horizon_mean_l2": 33.72223663330078, "reference_mean_abs_by_joint": { "shoulder_pan": 5.658662796020508, "shoulder_lift": 11.566729545593262, "elbow_flex": 18.132225036621094, "wrist_flex": 24.096614837646484, "wrist_roll": 6.287636756896973, "gripper": 0.24514202773571014 } }, "blue": { "reference_first_action_l2": 23.1048641204834, "reference_horizon_mean_l2": 20.655107498168945, "reference_mean_abs_by_joint": { "shoulder_pan": 2.229098081588745, "shoulder_lift": 9.639896392822266, "elbow_flex": 3.175830602645874, "wrist_flex": 17.395198822021484, "wrist_roll": 1.20304536819458, "gripper": 1.7563707828521729 } } }, "reference_horizon_l2_mean": 18.574945012728374 }, { "index": 146421, "episode_index": 352, "frame_index": 275, "phase": 0.5329457364341085, "bin_index": 53, "target_position": "right", "target_color": "green", "prompt_bucket": "direct_color", "task": "Put the banana in the green bowl.", "source_repo": "rslxcvg/banana_green", "source_episode": 18, "colors_by_position": "blue,red,green", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 14.94505524635315, "reference_decision_pan_range": 23.03296661376953, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "red", "blue", "green" ], "reference_decision_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "reference_first_l2_mean": 34.76588694254557, "reference_early_l2_mean": 38.33881123860677, "reference_horizon_pair_l2_mean": 50.437808990478516, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "red": -2.4175825119018555, "green": 8.659340858459473 }, "reference_state_pan_range": 11.076923370361328, "reference_state_l2_mean": 33.57865333557129 }, "color_to_position": { "blue": "left", "red": "center", "green": "right" }, "reference_indices_by_prompt": { "blue": 46620, "red": 93227, "green": 146421 }, "pan_by_prompt": { "red": 2.7821083068847656, "green": 18.574859619140625, "blue": -7.109592437744141 }, "pan_final_by_prompt": { "red": -5.241630554199219, "green": 34.19779968261719, "blue": -33.21858215332031 }, "shoulder_pan_range": 25.684452056884766, "shoulder_pan_final_range": 67.4163818359375, "pairwise_horizon_l2_mean": 34.18395805358887, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 6.352756023406982, "horizon_mean_l2": 31.224756240844727, "mean_abs_by_joint": { "shoulder_pan": 28.012327194213867, "shoulder_lift": 8.150910377502441, "elbow_flex": 8.809810638427734, "wrist_flex": 5.0152106285095215, "wrist_roll": 0.8113782405853271, "gripper": 0.059282705187797546 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 4.629342079162598, "horizon_mean_l2": 24.884395599365234, "mean_abs_by_joint": { "shoulder_pan": 18.10541534423828, "shoulder_lift": 9.876681327819824, "elbow_flex": 11.73587703704834, "wrist_flex": 4.731492519378662, "wrist_roll": 3.7875001430511475, "gripper": 0.13753587007522583 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 10.096911430358887, "horizon_mean_l2": 46.44272232055664, "mean_abs_by_joint": { "shoulder_pan": 46.11774444580078, "shoulder_lift": 1.7257713079452515, "elbow_flex": 2.9610366821289062, "wrist_flex": 0.28371837735176086, "wrist_roll": 3.843108892440796, "gripper": 0.1920759528875351 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 14.94505524635315, "reference_decision_pan_range": 23.03296661376953, "reference_final_pan_range": 64.79120826721191, "reference_first_pan_rank": [ "red", "blue", "green" ], "reference_decision_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "reference_first_l2_mean": 34.76588694254557, "reference_early_l2_mean": 38.33881123860677, "reference_horizon_pair_l2_mean": 50.437808990478516, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "red": -2.4175825119018555, "green": 8.659340858459473 }, "reference_state_pan_range": 11.076923370361328, "reference_state_l2_mean": 33.57865333557129 }, "reference_pan_final_by_prompt": { "blue": -4.483516693115234, "red": -2.1978023052215576, "green": 18.549449920654297 }, "reference_final_pan_by_prompt": { "blue": -30.41758155822754, "red": -6.593406677246094, "green": 34.373626708984375 }, "predicted_final_pan_rank": [ "blue", "red", "green" ], "predicted_pan_rank": [ "blue", "red", "green" ], "reference_pan_rank": [ "blue", "red", "green" ], "reference_final_pan_rank": [ "blue", "red", "green" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 23.4328670501709, "reference_horizon_mean_l2": 20.140960693359375, "reference_mean_abs_by_joint": { "shoulder_pan": 2.247666835784912, "shoulder_lift": 10.002055168151855, "elbow_flex": 4.47645378112793, "wrist_flex": 16.143484115600586, "wrist_roll": 1.028846263885498, "gripper": 1.618834137916565 } }, "green": { "reference_first_action_l2": 0.8698416352272034, "reference_horizon_mean_l2": 1.4049593210220337, "reference_mean_abs_by_joint": { "shoulder_pan": 0.3611999452114105, "shoulder_lift": 0.6394419074058533, "elbow_flex": 0.20110957324504852, "wrist_flex": 0.713275134563446, "wrist_roll": 0.7102332711219788, "gripper": 0.08113715797662735 } }, "blue": { "reference_first_action_l2": 44.862754821777344, "reference_horizon_mean_l2": 34.605995178222656, "reference_mean_abs_by_joint": { "shoulder_pan": 4.4279632568359375, "shoulder_lift": 12.349833488464355, "elbow_flex": 18.894672393798828, "wrist_flex": 24.39146614074707, "wrist_roll": 7.178341388702393, "gripper": 0.19664768874645233 } } }, "reference_horizon_l2_mean": 18.717305064201355 }, { "index": 146939, "episode_index": 353, "frame_index": 276, "phase": 0.5348837209302325, "bin_index": 53, "target_position": "right", "target_color": "red", "prompt_bucket": "direct_color", "task": "Put the banana in the red bowl.", "source_repo": "rslxcvg/banana_green", "source_episode": 18, "colors_by_position": "blue,green,red", "decision_score": 36.40541681647301, "is_decision_stage": true, "chunk_decision_overlap": 1.0, "valid_action_horizon": 30, "loss_weight": 2.0, "loss_weight_normalized": 1.5395125553914328, "sample_weight": 2.005934242181236, "reference_branch_selection": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "blue", "red" ], "reference_decision_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "green": -2.4175825119018555, "red": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "color_to_position": { "blue": "left", "green": "center", "red": "right" }, "reference_indices_by_prompt": { "blue": 47005, "green": 93653, "red": 146939 }, "pan_by_prompt": { "red": 19.933380126953125, "green": 1.0415077209472656, "blue": -10.272388458251953 }, "pan_final_by_prompt": { "red": 34.19779968261719, "green": -6.0270233154296875, "blue": -34.746910095214844 }, "shoulder_pan_range": 30.205768585205078, "shoulder_pan_final_range": 68.94470977783203, "pairwise_horizon_l2_mean": 36.5684445699056, "pairs": [ { "prompt_a": "red", "prompt_b": "green", "first_action_l2": 7.949377536773682, "horizon_mean_l2": 33.46841049194336, "mean_abs_by_joint": { "shoulder_pan": 30.62464141845703, "shoulder_lift": 7.404265403747559, "elbow_flex": 9.026355743408203, "wrist_flex": 4.839975833892822, "wrist_roll": 1.2282050848007202, "gripper": 0.08773843199014664 } }, { "prompt_a": "red", "prompt_b": "blue", "first_action_l2": 13.48166275024414, "horizon_mean_l2": 50.01468276977539, "mean_abs_by_joint": { "shoulder_pan": 49.34598922729492, "shoulder_lift": 2.123901128768921, "elbow_flex": 4.059236526489258, "wrist_flex": 0.756597638130188, "wrist_roll": 5.245832443237305, "gripper": 0.15650637447834015 } }, { "prompt_a": "green", "prompt_b": "blue", "first_action_l2": 6.4239373207092285, "horizon_mean_l2": 26.222240447998047, "mean_abs_by_joint": { "shoulder_pan": 18.721345901489258, "shoulder_lift": 9.528165817260742, "elbow_flex": 12.875771522521973, "wrist_flex": 4.083378791809082, "wrist_roll": 5.111217975616455, "gripper": 0.08773840218782425 } } ], "reference_branch_metrics": { "reference_action_index": 5, "reference_first_pan_range": 16.175824403762817, "reference_decision_pan_range": 25.23076868057251, "reference_final_pan_range": 65.49450492858887, "reference_first_pan_rank": [ "green", "blue", "red" ], "reference_decision_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "reference_first_l2_mean": 34.82464408874512, "reference_early_l2_mean": 38.58340962727865, "reference_horizon_pair_l2_mean": 50.88843790690104, "reference_state_pan_by_prompt": { "blue": -0.3956044018268585, "green": -2.4175825119018555, "red": 9.802197456359863 }, "reference_state_pan_range": 12.219779968261719, "reference_state_l2_mean": 33.8428757985433 }, "reference_pan_final_by_prompt": { "blue": -5.626373767852783, "green": -2.1978023052215576, "red": 19.604394912719727 }, "reference_final_pan_by_prompt": { "blue": -31.120878219604492, "green": -6.593406677246094, "red": 34.373626708984375 }, "predicted_final_pan_rank": [ "blue", "green", "red" ], "predicted_pan_rank": [ "blue", "green", "red" ], "reference_pan_rank": [ "blue", "green", "red" ], "reference_final_pan_rank": [ "blue", "green", "red" ], "rank_action_index": 5, "pan_rank_matches_reference": true, "reference_metrics_by_prompt": { "red": { "reference_first_action_l2": 1.4382072687149048, "reference_horizon_mean_l2": 1.5372731685638428, "reference_mean_abs_by_joint": { "shoulder_pan": 0.5307675004005432, "shoulder_lift": 0.5473600029945374, "elbow_flex": 0.1995185911655426, "wrist_flex": 0.5078887939453125, "wrist_roll": 0.8138737082481384, "gripper": 0.12144944071769714 } }, "green": { "reference_first_action_l2": 23.018178939819336, "reference_horizon_mean_l2": 20.41402244567871, "reference_mean_abs_by_joint": { "shoulder_pan": 2.335083246231079, "shoulder_lift": 10.664005279541016, "elbow_flex": 4.362128257751465, "wrist_flex": 16.104541778564453, "wrist_roll": 1.0714747905731201, "gripper": 1.6069777011871338 } }, "blue": { "reference_first_action_l2": 43.27715301513672, "reference_horizon_mean_l2": 32.90886688232422, "reference_mean_abs_by_joint": { "shoulder_pan": 5.529723167419434, "shoulder_lift": 12.147753715515137, "elbow_flex": 17.6275634765625, "wrist_flex": 23.36448097229004, "wrist_roll": 4.643772602081299, "gripper": 0.12995195388793945 } } }, "reference_horizon_l2_mean": 18.286720832188923 } ] }