{"url":"/sota/bench2drive-on-bench2drive","task":{"name":"Bench2Drive","url":"/task/bench2drive","note":null},"dataset":{"name":"Bench2Drive","url":null},"category":"Computer Vision","categories":["Computer Vision","Miscellaneous","Robots"],"category_note":null,"description":"Bench2Drive is an autonomous driving benchmark based on the CARLA leaderboard 2.0. It consists of 220 short routes featuring safety critical scenarios. The evaluation is performed closed-loop in the CARLA simulator. The performance of an entire driving stack is being evaluated.","description_from":"task","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":["Driving Score"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Driving Score":"higher"}},"counts":{"rows":35,"rows_with_code":20,"rows_with_paper_page":30,"rows_dated":30,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"HiP-AD","metrics":{"Driving Score":"86.77"},"uses_additional_data":false,"paper_date":"2025-03-11","paper":"/paper/hip-ad-hierarchical-and-multi-granularity","paper_url":"https://arxiv.org/abs/2503.08612v1","paper_title":"HiP-AD: Hierarchical and Multi-Granularity Planning with Deformable Attention for Autonomous Driving in a Single Decoder","code":"https://github.com/nullmax-vision/hip-ad","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"R2SE","metrics":{"Driving Score":"86.28"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":3,"model":"SimLingo-Base (CarLLaVa)","metrics":{"Driving Score":"85.94"},"uses_additional_data":false,"paper_date":"2024-06-14","paper":"/paper/carllava-vision-language-models-for-camera","paper_url":"https://arxiv.org/abs/2406.10165v1","paper_title":"CarLLaVA: Vision language models for camera-only closed-loop driving","code":"https://github.com/RenzKa/simlingo","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"TransFuser++","metrics":{"Driving Score":"84.21"},"uses_additional_data":false,"paper_date":"2023-06-13","paper":"/paper/hidden-biases-of-end-to-end-driving-models","paper_url":"https://arxiv.org/abs/2306.07957v2","paper_title":"Hidden Biases of End-to-End Driving Models","code":"https://github.com/autonomousvision/carla_garage","n_code_links":1,"syntology":{"n_ran":13,"n_unverified":4,"n_samples":17,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"GaussianFusion","metrics":{"Driving Score":"79.4"},"uses_additional_data":false,"paper_date":"2025-05-27","paper":"/paper/gaussianfusion-gaussian-based-multi-sensor","paper_url":"https://arxiv.org/abs/2506.00034v1","paper_title":"GaussianFusion: Gaussian-Based Multi-Sensor Fusion for End-to-End Autonomous Driving","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"ORION","metrics":{"Driving Score":"77.7"},"uses_additional_data":false,"paper_date":"2025-03-25","paper":"/paper/orion-a-holistic-end-to-end-autonomous","paper_url":"https://arxiv.org/abs/2503.19755v1","paper_title":"ORION: A Holistic End-to-End Autonomous Driving Framework by Vision-Language Instructed Action Generation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"Raw2Drive","metrics":{"Driving Score":"74.36"},"uses_additional_data":false,"paper_date":"2025-05-22","paper":"/paper/raw2drive-reinforcement-learning-with-aligned","paper_url":"https://arxiv.org/abs/2505.16394v1","paper_title":"Raw2Drive: Reinforcement Learning with Aligned World Models for End-to-End Autonomous Driving (in CARLA v2)","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":8,"model":"ETA","metrics":{"Driving Score":"74.33"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":9,"model":"DriveMoE","metrics":{"Driving Score":"74.22"},"uses_additional_data":false,"paper_date":"2025-05-22","paper":"/paper/drivemoe-mixture-of-experts-for-vision","paper_url":"https://arxiv.org/abs/2505.16278v1","paper_title":"DriveMoE: Mixture-of-Experts for Vision-Language-Action Model in End-to-End Autonomous Driving","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":10,"model":"Hydra-NeXt","metrics":{"Driving Score":"73.86"},"uses_additional_data":false,"paper_date":"2025-03-15","paper":"/paper/hydra-next-robust-closed-loop-driving-with","paper_url":"https://arxiv.org/abs/2503.12030v1","paper_title":"Hydra-NeXt: Robust Closed-Loop Driving with Open-Loop Training","code":"https://github.com/woxihuanjiangguo/hydra-next","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"VL (on failure)","metrics":{"Driving Score":"73.29"},"uses_additional_data":false,"paper_date":"2024-06-03","paper":"/paper/learning-from-mistakes-a-weakly-supervised","paper_url":"https://arxiv.org/abs/2406.01544v2","paper_title":"Validity Learning on Failures: Mitigating the Distribution Shift in Autonomous Vehicle Planning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":12,"model":"DRIVER","metrics":{"Driving Score":"68.90"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":13,"model":"DiffAD","metrics":{"Driving Score":"67.92"},"uses_additional_data":false,"paper_date":"2025-03-15","paper":"/paper/diffad-a-unified-diffusion-modeling-approach","paper_url":"https://arxiv.org/abs/2503.12170v1","paper_title":"DiffAD: A Unified Diffusion Modeling Approach for Autonomous Driving","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":14,"model":"NavigationDrive","metrics":{"Driving Score":"67.17"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":15,"model":"iPad","metrics":{"Driving Score":"65.02"},"uses_additional_data":false,"paper_date":"2025-05-21","paper":"/paper/ipad-iterative-proposal-centric-end-to-end","paper_url":"https://arxiv.org/abs/2505.15111v1","paper_title":"iPad: Iterative Proposal-centric End-to-End Autonomous Driving","code":"https://github.com/Kguo-cs/iPad","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"DriveAdapter","metrics":{"Driving Score":"64.22"},"uses_additional_data":false,"paper_date":"2023-08-01","paper":"/paper/driveadapter-breaking-the-coupling-barrier-of","paper_url":"https://arxiv.org/abs/2308.00398v2","paper_title":"DriveAdapter: Breaking the Coupling Barrier of Perception and Planning in End-to-End Autonomous Driving","code":"https://github.com/opendrivelab/driveadapter","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":1,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":17,"model":"ReasonPlan","metrics":{"Driving Score":"64.01"},"uses_additional_data":false,"paper_date":"2025-05-26","paper":"/paper/reasonplan-unified-scene-prediction-and","paper_url":"https://arxiv.org/abs/2505.20024v1","paper_title":"ReasonPlan: Unified Scene Prediction and Decision Reasoning for Closed-loop Autonomous Driving","code":"https://github.com/liuxueyi/reasonplan","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"Drivetransformer-Large","metrics":{"Driving Score":"63.46"},"uses_additional_data":false,"paper_date":"2025-03-07","paper":"/paper/drivetransformer-unified-transformer-for","paper_url":"https://arxiv.org/abs/2503.07656v1","paper_title":"DriveTransformer: Unified Transformer for Scalable End-to-End Autonomous Driving","code":"https://github.com/thinklab-sjtu/drivetransformer","n_code_links":1,"syntology":null},{"rank_in_archive_order":19,"model":"ThinkTwice","metrics":{"Driving Score":"62.44"},"uses_additional_data":false,"paper_date":"2023-05-10","paper":"/paper/think-twice-before-driving-towards-scalable","paper_url":"https://arxiv.org/abs/2305.06242v1","paper_title":"Think Twice before Driving: Towards Scalable Decoders for End-to-End Autonomous Driving","code":"https://github.com/opendrivelab/thinktwice","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":3,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":20,"model":"TCP-traj","metrics":{"Driving Score":"59.90"},"uses_additional_data":false,"paper_date":"2022-06-16","paper":"/paper/trajectory-guided-control-prediction-for-end","paper_url":"https://arxiv.org/abs/2206.08129v2","paper_title":"Trajectory-guided Control Prediction for End-to-end Autonomous Driving: A Simple yet Strong Baseline","code":"https://github.com/OpenPerceptionX/TCP","n_code_links":1,"syntology":null},{"rank_in_archive_order":21,"model":"DiFSD","metrics":{"Driving Score":"52.02"},"uses_additional_data":false,"paper_date":"2024-09-15","paper":"/paper/difsd-ego-centric-fully-sparse-paradigm-with","paper_url":"https://arxiv.org/abs/2409.09777v4","paper_title":"DiFSD: Ego-Centric Fully Sparse Paradigm with Uncertainty Denoising and Iterative Refinement for Efficient End-to-End Self-Driving","code":"https://github.com/suhaisheng/difsd","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"X-Driver","metrics":{"Driving Score":"51.70"},"uses_additional_data":false,"paper_date":"2025-05-08","paper":"/paper/x-driver-explainable-autonomous-driving-with","paper_url":"https://arxiv.org/abs/2505.05098v2","paper_title":"X-Driver: Explainable Autonomous Driving with Vision-Language Models","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":23,"model":"TCP-traj w/o distillation","metrics":{"Driving Score":"49.30"},"uses_additional_data":false,"paper_date":"2022-06-16","paper":"/paper/trajectory-guided-control-prediction-for-end","paper_url":"https://arxiv.org/abs/2206.08129v2","paper_title":"Trajectory-guided Control Prediction for End-to-end Autonomous Driving: A Simple yet Strong Baseline","code":"https://github.com/OpenPerceptionX/TCP","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"CogAD","metrics":{"Driving Score":"48.30"},"uses_additional_data":false,"paper_date":"2025-05-27","paper":"/paper/cogad-cognitive-hierarchy-guided-end-to-end","paper_url":"https://arxiv.org/abs/2505.21581v2","paper_title":"CogAD: Cognitive-Hierarchy Guided End-to-End Autonomous Driving","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":25,"model":"MomAD","metrics":{"Driving Score":"47.91"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":26,"model":"UniAD-Base","metrics":{"Driving Score":"45.81"},"uses_additional_data":false,"paper_date":"2022-12-20","paper":"/paper/goal-oriented-autonomous-driving","paper_url":"https://arxiv.org/abs/2212.10156v2","paper_title":"Planning-oriented Autonomous Driving","code":"https://github.com/opendrivelab/uniad","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":27,"model":"TTOG","metrics":{"Driving Score":"45.23"},"uses_additional_data":false,"paper_date":"2025-04-17","paper":"/paper/two-tasks-one-goal-uniting-motion-and","paper_url":"https://arxiv.org/abs/2504.12667v1","paper_title":"Two Tasks, One Goal: Uniting Motion and Planning for Excellent End To End Autonomous Driving Performance","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":28,"model":"GenAD","metrics":{"Driving Score":"44.81"},"uses_additional_data":false,"paper_date":"2024-02-18","paper":"/paper/genad-generative-end-to-end-autonomous","paper_url":"https://arxiv.org/abs/2402.11502v3","paper_title":"GenAD: Generative End-to-End Autonomous Driving","code":"https://github.com/wzzheng/genad","n_code_links":1,"syntology":null},{"rank_in_archive_order":29,"model":"SparseDrive","metrics":{"Driving Score":"44.54"},"uses_additional_data":false,"paper_date":"2024-05-30","paper":"/paper/sparsedrive-end-to-end-autonomous-driving-via","paper_url":"https://arxiv.org/abs/2405.19620v2","paper_title":"SparseDrive: End-to-End Autonomous Driving via Sparse Scene Representation","code":"https://github.com/swc-17/sparsedrive","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":30,"model":"VAD","metrics":{"Driving Score":"42.35"},"uses_additional_data":false,"paper_date":"2023-03-21","paper":"/paper/vad-vectorized-scene-representation-for","paper_url":"https://arxiv.org/abs/2303.12077v3","paper_title":"VAD: Vectorized Scene Representation for Efficient Autonomous Driving","code":"https://github.com/hustvl/vad","n_code_links":2,"syntology":null},{"rank_in_archive_order":31,"model":"UniAD-Tiny","metrics":{"Driving Score":"40.73"},"uses_additional_data":false,"paper_date":"2022-12-20","paper":"/paper/goal-oriented-autonomous-driving","paper_url":"https://arxiv.org/abs/2212.10156v2","paper_title":"Planning-oriented Autonomous Driving","code":"https://github.com/opendrivelab/uniad","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":32,"model":"TCP","metrics":{"Driving Score":"40.70"},"uses_additional_data":false,"paper_date":"2022-06-16","paper":"/paper/trajectory-guided-control-prediction-for-end","paper_url":"https://arxiv.org/abs/2206.08129v2","paper_title":"Trajectory-guided Control Prediction for End-to-end Autonomous Driving: A Simple yet Strong Baseline","code":"https://github.com/OpenPerceptionX/TCP","n_code_links":1,"syntology":null},{"rank_in_archive_order":33,"model":"VAD + SERA","metrics":{"Driving Score":"35.64"},"uses_additional_data":false,"paper_date":"2025-05-28","paper":"/paper/from-failures-to-fixes-llm-driven-scenario","paper_url":"https://arxiv.org/abs/2505.22067v1","paper_title":"From Failures to Fixes: LLM-Driven Scenario Repair for Self-Evolving Autonomous Driving","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":34,"model":"TCP-ctrl","metrics":{"Driving Score":"30.47"},"uses_additional_data":false,"paper_date":"2022-06-16","paper":"/paper/trajectory-guided-control-prediction-for-end","paper_url":"https://arxiv.org/abs/2206.08129v2","paper_title":"Trajectory-guided Control Prediction for End-to-end Autonomous Driving: A Simple yet Strong Baseline","code":"https://github.com/OpenPerceptionX/TCP","n_code_links":1,"syntology":null},{"rank_in_archive_order":35,"model":"AD-MLP","metrics":{"Driving Score":"18.05"},"uses_additional_data":false,"paper_date":"2024-06-06","paper":"/paper/bench2drive-towards-multi-ability","paper_url":"https://arxiv.org/abs/2406.03877v3","paper_title":"Bench2Drive: Towards Multi-Ability Benchmarking of Closed-Loop End-To-End Autonomous Driving","code":"https://github.com/Thinklab-SJTU/Bench2Drive","n_code_links":4,"syntology":{"n_ran":13,"n_unverified":4,"n_samples":17,"n_pointer_only_licence":17}}],"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":7,"rows_with_any_sample_ran":6,"distinct_papers_with_graph_line":6,"distinct_papers_with_any_sample_ran":5,"samples_over_distinct_papers":{"n_ran":37,"n_unverified":14,"n_samples":51,"n_pointer_only_licence":17,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":38,"n_unverified":14,"n_samples":52,"n_pointer_only_licence":17,"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"}}}