{"paper":{"title":"Causality-Aware End-to-End Autonomous Driving via Ego-Centric Joint Scene Modeling","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"CaAD models causal dependencies between the ego vehicle and surrounding agents inside a shared latent scene representation to produce more consistent closed-loop trajectories.","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Jinkyu Kim, Joon Seo, Jungbeom Lee, Minseung Lee, Seokha Moon","submitted_at":"2026-05-13T15:06:22Z","abstract_excerpt":"End-to-end autonomous driving, which bypasses traditional modular pipelines by directly predicting future trajectories from sensor inputs, has recently achieved substantial progress. However, existing methods often overlook the causal inter-dependencies in ego-vehicle planning, ignoring the reciprocal relations between the ego vehicle and surrounding agents. This causal oversight leads to inconsistent and unreliable trajectory predictions, especially in interaction-critical scenarios where ego decisions and neighboring agent behaviors must be reasoned about jointly. To address this limitation,"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"we propose CaAD, a Causality-aware end-to-end Autonomous Driving framework that captures these dependencies within a shared latent scene representation. On the Bench2Drive and NAVSIM benchmarks, CaAD demonstrates strong closed-loop planning performance, achieving a Driving Score of 87.53 and Success Rate of 71.81 on Bench2Drive, and a PDMS of 91.1 on NAVSIM.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the ego-centric joint-causal modeling module actually learns genuine causal dependencies (rather than correlations) between the ego vehicle and interaction-relevant agents, and that aligning the stochastic ego policy via joint-mode embeddings will produce more consistent and reliable closed-loop trajectories in interaction-critical scenarios.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CaAD adds ego-centric joint-causal modeling and causality-aware policy alignment to end-to-end driving, reporting Driving Score 87.53 and Success Rate 71.81 on Bench2Drive plus PDMS 91.1 on NAVSIM.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"CaAD models causal dependencies between the ego vehicle and surrounding agents inside a shared latent scene representation to produce more consistent closed-loop trajectories.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"9a2ef0b4b46eaf7bf9974c72512c910cd5233cb5d5097176f4a40ed57bd2378c"},"source":{"id":"2605.13646","kind":"arxiv","version":1},"verdict":{"id":"aafa4a76-3ffa-4041-82f1-57ad60fe8920","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T18:21:52.857565Z","strongest_claim":"we propose CaAD, a Causality-aware end-to-end Autonomous Driving framework that captures these dependencies within a shared latent scene representation. On the Bench2Drive and NAVSIM benchmarks, CaAD demonstrates strong closed-loop planning performance, achieving a Driving Score of 87.53 and Success Rate of 71.81 on Bench2Drive, and a PDMS of 91.1 on NAVSIM.","one_line_summary":"CaAD adds ego-centric joint-causal modeling and causality-aware policy alignment to end-to-end driving, reporting Driving Score 87.53 and Success Rate 71.81 on Bench2Drive plus PDMS 91.1 on NAVSIM.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the ego-centric joint-causal modeling module actually learns genuine causal dependencies (rather than correlations) between the ego vehicle and interaction-relevant agents, and that aligning the stochastic ego policy via joint-mode embeddings will produce more consistent and reliable closed-loop trajectories in interaction-critical scenarios.","pith_extraction_headline":"CaAD models causal dependencies between the ego vehicle and surrounding agents inside a shared latent scene representation to produce more consistent closed-loop trajectories."},"references":{"count":66,"sample":[{"doi":"","year":2020,"title":"nuscenes: A multimodal dataset for autonomous driving","work_id":"2c148a70-8b96-49e6-b0fe-0566eac37bd0","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"NuPlan: A closed-loop ML-based planning benchmark for autonomous vehicles","work_id":"b16aace6-deff-4546-b333-bcb7c9c07cdb","ref_index":2,"cited_arxiv_id":"2106.11810","is_internal_anchor":true},{"doi":"","year":2023,"title":"Gri: General reinforced imitation and its application to vision-based autonomous driving.Robotics, 12(5):127, 2023","work_id":"1055d269-8286-40ae-9f69-93c45f0cdebd","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"End-to-end autonomous driving: Challenges and frontiers.IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(12):10164–10183, 2024","work_id":"87741bb5-68f9-46cc-bc5b-c7f27777562d","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"VADv2: End-to-End Vectorized Autonomous Driving via Probabilistic Planning","work_id":"e7670f83-e1e1-41e7-86eb-39477a3a10b2","ref_index":5,"cited_arxiv_id":"2402.13243","is_internal_anchor":true}],"resolved_work":66,"snapshot_sha256":"4b963d0be4f216096951f691c49f892ced149b863022da8605a1a95c61ff732e","internal_anchors":13},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}