{"url":"/sota/robot-manipulation-on-rlbench","task":{"name":"Robot Manipulation","url":"/task/robot-manipulation","note":null},"dataset":{"name":"RLBench","url":"/dataset/rlbench"},"category":"Robots","categories":["Robots"],"category_note":null,"description":null,"description_from":null,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["Succ. Rate (18 tasks, 100 demo/task)","Succ. Rate (18 tasks, 10 demo/task)","Training Time (V100 x 8 x day)","Training Time (A100 x hour)","Succ. Rate (10 tasks, 100 demos/task)","Succ. Rate (74 tasks, 100 demos/task)","Inference Speed (fps)","Input Image Size"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Succ. Rate (18 tasks, 100 demo/task)":null,"Succ. Rate (18 tasks, 10 demo/task)":null,"Training Time (V100 x 8 x day)":"lower","Training Time (A100 x hour)":"lower","Succ. Rate (10 tasks, 100 demos/task)":null,"Succ. Rate (74 tasks, 100 demos/task)":null,"Inference Speed (fps)":null,"Input Image Size":null}},"counts":{"rows":18,"rows_with_code":16,"rows_with_paper_page":17,"rows_dated":18,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"EquAct","metrics":{"Inference Speed (fps)":"1.4","Input Image Size":"256","Succ. Rate (18 tasks, 10 demo/task)":"60.1","Succ. Rate (18 tasks, 100 demo/task)":"89.4","Training Time (A100 x hour)":"240"},"uses_additional_data":false,"paper_date":"2025-05-27","paper":null,"paper_url":"https://arxiv.org/abs/2505.21351v1","paper_title":"EquAct: An SE(3)-Equivariant Multi-Task Transformer for Open-Loop Robotic Manipulation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":2,"model":"SAM2Act","metrics":{"Input Image Size":"128","Succ. Rate (18 tasks, 100 demo/task)":"86.8","Training Time (A100 x hour)":"200"},"uses_additional_data":false,"paper_date":"2025-01-30","paper":"/paper/sam2act-integrating-visual-foundation-model-1","paper_url":"https://arxiv.org/abs/2501.18564v3","paper_title":"SAM2Act: Integrating Visual Foundation Model with A Memory Architecture for Robotic Manipulation","code":"https://github.com/sam2act/sam2act","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"ARP+","metrics":{"Succ. Rate (18 tasks, 100 demo/task)":"84.9"},"uses_additional_data":false,"paper_date":"2024-10-04","paper":"/paper/autoregressive-action-sequence-learning-for","paper_url":"https://arxiv.org/abs/2410.03132v5","paper_title":"Autoregressive Action Sequence Learning for Robotic Manipulation","code":"https://github.com/mlzxy/arp","n_code_links":1,"syntology":{"n_ran":13,"n_unverified":1,"n_samples":14,"n_pointer_only_licence":14}},{"rank_in_archive_order":4,"model":"3D-LOTUS","metrics":{"Inference Speed (fps)":"9.5","Input Image Size":"256","Succ. Rate (18 tasks, 100 demo/task)":"83.1","Training Time (A100 x hour)":"40","Training Time (V100 x 8 x day)":"0.28"},"uses_additional_data":false,"paper_date":"2024-10-02","paper":"/paper/towards-generalizable-vision-language-robotic","paper_url":"https://arxiv.org/abs/2410.01345v2","paper_title":"Towards Generalizable Vision-Language Robotic Manipulation: A Benchmark and LLM-guided 3D Policy","code":"https://github.com/vlc-robot/robot-3dlotus","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":5,"model":"RVT-2","metrics":{"Inference Speed (fps)":"20.6","Input Image Size":"128","Succ. Rate (18 tasks, 100 demo/task)":"81.4","Training Time (V100 x 8 x day)":"0.83"},"uses_additional_data":false,"paper_date":"2024-06-12","paper":"/paper/rvt-2-learning-precise-manipulation-from-few","paper_url":"https://arxiv.org/abs/2406.08545v1","paper_title":"RVT-2: Learning Precise Manipulation from Few Demonstrations","code":"https://github.com/NVlabs/RVT","n_code_links":1,"syntology":{"n_ran":15,"n_unverified":0,"n_samples":15,"n_pointer_only_licence":15}},{"rank_in_archive_order":6,"model":"3D Diffuser Actor","metrics":{"Input Image Size":"256","Succ. Rate (18 tasks, 100 demo/task)":"81.3","Training Time (A100 x hour)":"936","Training Time (V100 x 8 x day)":"8"},"uses_additional_data":false,"paper_date":"2024-02-18","paper":"/paper/3d-diffuser-actor-policy-diffusion-with-3d","paper_url":"https://arxiv.org/abs/2402.10885","paper_title":"3D Diffuser Actor: Policy Diffusion with 3D Scene Representations","code":"https://github.com/nickgkan/3d_diffuser_actor","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":9,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"Mini Diffuser","metrics":{"Input Image Size":"256","Succ. Rate (18 tasks, 100 demo/task)":"77.6","Training Time (A100 x hour)":"24","Training Time (V100 x 8 x day)":"0.24"},"uses_additional_data":false,"paper_date":"2025-05-14","paper":"/paper/train-a-multi-task-diffusion-policy-on","paper_url":"https://arxiv.org/abs/2505.09430v2","paper_title":"Mini Diffuser: Fast Multi-task Diffusion Policy Training Using Two-level Mini-batches","code":"https://github.com/utomm/mini-diffuse-actor","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"SAM-E","metrics":{"Succ. Rate (18 tasks, 100 demo/task)":"70.6"},"uses_additional_data":false,"paper_date":"2024-05-30","paper":"/paper/sam-e-leveraging-visual-foundation-model-with","paper_url":"https://arxiv.org/abs/2405.19586v1","paper_title":"SAM-E: Leveraging Visual Foundation Model with Sequence Imitation for Embodied Manipulation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":9,"model":"Act3D","metrics":{"Input Image Size":"256","Succ. Rate (18 tasks, 10 demo/task)":"48","Succ. Rate (18 tasks, 100 demo/task)":"65","Training Time (V100 x 8 x day)":"5"},"uses_additional_data":false,"paper_date":"2023-06-30","paper":"/paper/act3d-infinite-resolution-action-detection","paper_url":"https://arxiv.org/abs/2306.17817v2","paper_title":"Act3D: 3D Feature Field Transformers for Multi-Task Robotic Manipulation","code":"https://github.com/nickgkan/3d_diffuser_actor","n_code_links":2,"syntology":null},{"rank_in_archive_order":10,"model":"RVT","metrics":{"Inference Speed (fps)":"11.6","Input Image Size":"128","Succ. Rate (18 tasks, 100 demo/task)":"62.9","Training Time (V100 x 8 x day)":"1"},"uses_additional_data":false,"paper_date":"2023-06-26","paper":"/paper/rvt-robotic-view-transformer-for-3d-object","paper_url":"https://arxiv.org/abs/2306.14896v1","paper_title":"RVT: Robotic View Transformer for 3D Object Manipulation","code":"https://github.com/NVlabs/RVT","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"PerAct (Evaluated in RVT)","metrics":{"Inference Speed (fps)":"4.9","Input Image Size":"128","Succ. Rate (18 tasks, 100 demo/task)":"49.4","Training Time (V100 x 8 x day)":"16"},"uses_additional_data":false,"paper_date":"2022-09-12","paper":"/paper/perceiver-actor-a-multi-task-transformer-for","paper_url":"https://arxiv.org/abs/2209.05451v2","paper_title":"Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation","code":"https://github.com/peract/peract","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"PolarNet","metrics":{"Input Image Size":"128","Succ. Rate (10 tasks, 100 demos/task)":"89.8","Succ. Rate (18 tasks, 100 demo/task)":"46.4","Succ. Rate (74 tasks, 100 demos/task)":"60.3","Training Time (V100 x 8 x day)":"5"},"uses_additional_data":false,"paper_date":"2023-09-27","paper":"/paper/polarnet-3d-point-clouds-for-language-guided","paper_url":"https://arxiv.org/abs/2309.15596v1","paper_title":"PolarNet: 3D Point Clouds for Language-Guided Robotic Manipulation","code":"https://github.com/vlc-robot/polarnet","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"Hiveformer","metrics":{"Succ. Rate (10 tasks, 100 demos/task)":"83.3","Succ. Rate (18 tasks, 100 demo/task)":"45.3"},"uses_additional_data":false,"paper_date":"2022-09-11","paper":"/paper/instruction-driven-history-aware-policies-for","paper_url":"https://arxiv.org/abs/2209.04899v3","paper_title":"Instruction-driven history-aware policies for robotic manipulations","code":"https://github.com/guhur/hiveformer","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":11,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"PerAct","metrics":{"Input Image Size":"128","Succ. Rate (18 tasks, 10 demo/task)":"30","Succ. Rate (18 tasks, 100 demo/task)":"42.7","Training Time (V100 x 8 x day)":"16"},"uses_additional_data":false,"paper_date":"2022-09-12","paper":"/paper/perceiver-actor-a-multi-task-transformer-for","paper_url":"https://arxiv.org/abs/2209.05451v2","paper_title":"Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation","code":"https://github.com/peract/peract","n_code_links":1,"syntology":null},{"rank_in_archive_order":15,"model":"C2FARM-BC (Evaluated in PerAct)","metrics":{"Input Image Size":"128","Succ. Rate (18 tasks, 100 demo/task)":"20.1"},"uses_additional_data":false,"paper_date":"2021-06-23","paper":"/paper/coarse-to-fine-q-attention-efficient-learning","paper_url":"https://arxiv.org/abs/2106.12534v2","paper_title":"Coarse-to-Fine Q-attention: Efficient Learning for Visual Robotic Manipulation via Discretisation","code":"https://github.com/stepjam/ARM","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"Image-BC VIT","metrics":{"Input Image Size":"128","Succ. Rate (18 tasks, 100 demo/task)":"1.3"},"uses_additional_data":false,"paper_date":"2022-09-12","paper":"/paper/perceiver-actor-a-multi-task-transformer-for","paper_url":"https://arxiv.org/abs/2209.05451v2","paper_title":"Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation","code":"https://github.com/peract/peract","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"Image-BC CNN","metrics":{"Input Image Size":"128","Succ. Rate (18 tasks, 100 demo/task)":"1.3"},"uses_additional_data":false,"paper_date":"2022-09-12","paper":"/paper/perceiver-actor-a-multi-task-transformer-for","paper_url":"https://arxiv.org/abs/2209.05451v2","paper_title":"Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation","code":"https://github.com/peract/peract","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"Auto-λ","metrics":{"Succ. Rate (10 tasks, 100 demos/task)":"69.3"},"uses_additional_data":false,"paper_date":"2022-02-07","paper":"/paper/auto-lambda-disentangling-dynamic-task","paper_url":"https://arxiv.org/abs/2202.03091v2","paper_title":"Auto-Lambda: Disentangling Dynamic Task Relationships","code":"https://github.com/lorenmt/auto-lambda","n_code_links":1,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":6,"rows_with_any_sample_ran":4,"distinct_papers_with_graph_line":6,"distinct_papers_with_any_sample_ran":4,"samples_over_distinct_papers":{"n_ran":30,"n_unverified":22,"n_samples":52,"n_pointer_only_licence":30,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":30,"n_unverified":22,"n_samples":52,"n_pointer_only_licence":30,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}