{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:W4TMZISDYX6ORWOXVNOFXJHHEP","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"17bd67d47ad2d196eae1dc38fc521caafe3df3643ca962a72a09347e804e71df","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.RO","submitted_at":"2026-04-30T16:59:07Z","title_canon_sha256":"5c67a6ebe7eabd6892c9172c34f806be3108bda550ebad88c37707cbfdc928c8"},"schema_version":"1.0","source":{"id":"2604.28111","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.28111","created_at":"2026-05-20T00:00:40Z"},{"alias_kind":"arxiv_version","alias_value":"2604.28111v3","created_at":"2026-05-20T00:00:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.28111","created_at":"2026-05-20T00:00:40Z"},{"alias_kind":"pith_short_12","alias_value":"W4TMZISDYX6O","created_at":"2026-05-20T00:00:40Z"},{"alias_kind":"pith_short_16","alias_value":"W4TMZISDYX6ORWOX","created_at":"2026-05-20T00:00:40Z"},{"alias_kind":"pith_short_8","alias_value":"W4TMZISD","created_at":"2026-05-20T00:00:40Z"}],"graph_snapshots":[{"event_id":"sha256:d7518ee24dfec3ff346dc344dde501e9b0205bd25cd879d80be27774790d74f1","target":"graph","created_at":"2026-05-20T00:00:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Evaluated on the reconstructed nuScenes dataset, our method outperforms other simulation-based RL approaches in closed-loop experiments."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The 3D Gaussian Splatting environment provides sufficiently accurate and differentiable simulation of future vehicle dynamics and interactions to produce useful dense shaping rewards that transfer to real-world policy improvement."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"GSDrive combines IL priors with RL feedback by probing multi-mode futures inside a 3D Gaussian Splatting simulator to supply dense rewards for closed-loop driving policy improvement on nuScenes."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A 3D Gaussian Splatting environment probes multiple candidate futures to supply dense rewards that refine end-to-end driving policies."}],"snapshot_sha256":"733cf257172b3189b6b2a1bf47a1fe2820881f329581bee168d367188accaf94"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"887639c3c704d66af01e9d9af0ab3c689fff7cf1dd88cee7710542bcdf464dec"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T18:36:41.646637Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2604.28111/integrity.json","findings":[],"snapshot_sha256":"16646272ba79230c2abdc934ac68516d519d0c6b26c78d656db7c469c766f2d2","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"End-to-end (E2E) autonomous driving aims to directly map sensory observations to driving actions, but its real-world deployment is hindered by evolving data distributions and the high cost of continual annotation. While combining imitation learning (IL) and reinforcement learning (RL) is a common strategy for policy improvement, conventional RL training relies on delayed, event-based rewards, where policies learn only from catastrophic outcomes such as collisions, leading to premature convergence to suboptimal behaviors. To address these limitations, we propose GSDrive, a framework that uses a","authors_text":"Chen Min, Dzmitry Tsetserukou, Shuo Wang, Sifa Zheng, Xuefeng Zhang, Yixiao Zhou, Ziang Guo, Zufeng Zhang","cross_cats":[],"headline":"A 3D Gaussian Splatting environment probes multiple candidate futures to supply dense rewards that refine end-to-end driving policies.","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.RO","submitted_at":"2026-04-30T16:59:07Z","title":"GSDrive: Reinforcing Driving Policies by Multi-mode Future Trajectory Probing with 3D Gaussian Splatting Environment"},"references":{"count":31,"internal_anchors":6,"resolved_work":31,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"End- to-end autonomous driving: Challenges and frontiers,","work_id":"e86bcc53-963a-47a3-bbfe-dd58108914ba","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"The era of end-to-end autonomy: Transitioning from rule-based driving to large driving models","work_id":"e401a554-5c4f-4b87-8aac-2f54e0f56118","year":2026},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Iterative label refinement matters more than preference optimization under weak supervision","work_id":"12bd644a-8578-4b18-8fa9-494b3d751a62","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"End-to-end driving with online trajectory evaluation via bev world model,","work_id":"9c7b66de-f25b-4880-b41d-6a56394155d5","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"arXiv preprint arXiv:2503.11650 (2025)","work_id":"d764e87f-6ac5-4436-bca1-d2eca7a1d8a0","year":2025}],"snapshot_sha256":"f50a50aaab8b82c3819e281ee017a4604326c69858ce4f187bb4a22bd1c18af8"},"source":{"id":"2604.28111","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-19T16:42:42.310490Z","id":"8452bb40-5f93-41a3-ac71-fdaa442894dd","model_set":{"reader":"grok-4.3"},"one_line_summary":"GSDrive combines IL priors with RL feedback by probing multi-mode futures inside a 3D Gaussian Splatting simulator to supply dense rewards for closed-loop driving policy improvement on nuScenes.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A 3D Gaussian Splatting environment probes multiple candidate futures to supply dense rewards that refine end-to-end driving policies.","strongest_claim":"Evaluated on the reconstructed nuScenes dataset, our method outperforms other simulation-based RL approaches in closed-loop experiments.","weakest_assumption":"The 3D Gaussian Splatting environment provides sufficiently accurate and differentiable simulation of future vehicle dynamics and interactions to produce useful dense shaping rewards that transfer to real-world policy improvement."}},"verdict_id":"8452bb40-5f93-41a3-ac71-fdaa442894dd"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d14a2bf825dc2ead9482aaa8afd2c4a73fda39736130264e93d15f80bed27316","target":"record","created_at":"2026-05-20T00:00:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"17bd67d47ad2d196eae1dc38fc521caafe3df3643ca962a72a09347e804e71df","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.RO","submitted_at":"2026-04-30T16:59:07Z","title_canon_sha256":"5c67a6ebe7eabd6892c9172c34f806be3108bda550ebad88c37707cbfdc928c8"},"schema_version":"1.0","source":{"id":"2604.28111","kind":"arxiv","version":3}},"canonical_sha256":"b726cca243c5fce8d9d7ab5c5ba4e723ef79f6f947300bfe7710be410e11f9aa","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b726cca243c5fce8d9d7ab5c5ba4e723ef79f6f947300bfe7710be410e11f9aa","first_computed_at":"2026-05-20T00:00:40.020230Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:00:40.020230Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lA+sNLgrfjAju9U65BLA/WT06b6wW9e4YGDE/kpdFWucyMSbGbOMM5+wHRE7L8L4SCyBr4PgyE/8SBhpX8rQAQ==","signature_status":"signed_v1","signed_at":"2026-05-20T00:00:40.021014Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.28111","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d14a2bf825dc2ead9482aaa8afd2c4a73fda39736130264e93d15f80bed27316","sha256:d7518ee24dfec3ff346dc344dde501e9b0205bd25cd879d80be27774790d74f1"],"state_sha256":"a5f85935b017be242dca247cd62d644dda55aad573583481ebac30af39f92a5d"}