{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2SZANLNDCKHQJ5JQCZJTLIDITD","short_pith_number":"pith:2SZANLND","schema_version":"1.0","canonical_sha256":"d4b206ada3128f04f530165335a06898ff967a889e088fc497082cdd43215b92","source":{"kind":"arxiv","id":"2509.07953","version":1},"attestation_state":"computed","paper":{"title":"RaC: Robot Learning for Long-Horizon Tasks by Scaling Recovery and Correction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Aviral Kumar, Jasmine Li, Naveen Enock, Riya Kadakia, Robyn Wu, Zackory Erickson, Zheyuan Hu","submitted_at":"2025-09-09T17:41:29Z","abstract_excerpt":"Modern paradigms for robot imitation train expressive policy architectures on large amounts of human demonstration data. Yet performance on contact-rich, deformable-object, and long-horizon tasks plateau far below perfect execution, even with thousands of expert demonstrations. This is due to the inefficiency of existing ``expert'' data collection procedures based on human teleoperation. To address this issue, we introduce RaC, a new phase of training on human-in-the-loop rollouts after imitation learning pre-training. In RaC, we fine-tune a robotic policy on human intervention trajectories th"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2509.07953","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-09-09T17:41:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"74d77e984d0936ea19350ab4616387982ca01f5593a27d25eac744925c2826a7","abstract_canon_sha256":"3dfc649ca12fd1006874d34333831a6d4e52e8ed289610a45b9e5ce79a334f0f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:07:42.958252Z","signature_b64":"QupB/p2B6k/hcoRGJqedXsqO2ZjNFkQG0++EgXdQr9rlDt6o2C/GKsOoK1Q+33rX0Hg/LItDxSq4UTjAwrGmCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d4b206ada3128f04f530165335a06898ff967a889e088fc497082cdd43215b92","last_reissued_at":"2026-07-05T12:07:42.957699Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:07:42.957699Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RaC: Robot Learning for Long-Horizon Tasks by Scaling Recovery and Correction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Aviral Kumar, Jasmine Li, Naveen Enock, Riya Kadakia, Robyn Wu, Zackory Erickson, Zheyuan Hu","submitted_at":"2025-09-09T17:41:29Z","abstract_excerpt":"Modern paradigms for robot imitation train expressive policy architectures on large amounts of human demonstration data. Yet performance on contact-rich, deformable-object, and long-horizon tasks plateau far below perfect execution, even with thousands of expert demonstrations. This is due to the inefficiency of existing ``expert'' data collection procedures based on human teleoperation. To address this issue, we introduce RaC, a new phase of training on human-in-the-loop rollouts after imitation learning pre-training. In RaC, we fine-tune a robotic policy on human intervention trajectories th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.07953","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2509.07953/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2509.07953","created_at":"2026-07-05T12:07:42.957758+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.07953v1","created_at":"2026-07-05T12:07:42.957758+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.07953","created_at":"2026-07-05T12:07:42.957758+00:00"},{"alias_kind":"pith_short_12","alias_value":"2SZANLNDCKHQ","created_at":"2026-07-05T12:07:42.957758+00:00"},{"alias_kind":"pith_short_16","alias_value":"2SZANLNDCKHQJ5JQ","created_at":"2026-07-05T12:07:42.957758+00:00"},{"alias_kind":"pith_short_8","alias_value":"2SZANLND","created_at":"2026-07-05T12:07:42.957758+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01804","citing_title":"VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09615","citing_title":"DexPIE: Stable Dexterous Policy Improvement from Real-World Experience","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01111","citing_title":"FAR: Failure-Aware Retry for Test-Time Recovery and Continual Policy Improvement","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01865","citing_title":"Set-Supervised Diffusion Policy: Learning Action-Chunking Diffusion through Corrections","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31286","citing_title":"DeMaVLA: A Vision-Language-Action Foundation Model for Generalizable Deformable Manipulation","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28939","citing_title":"ReGuide: From Test-Time Guidance to Self-Improving Diffusion Policies","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15157","citing_title":"Hand-in-the-Loop: Improving VLA Policies for Dexterous Manipulation via Seamless Hand-Arm Intervention","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2602.11075","citing_title":"RISE: Self-Improving Robot Policy with Compositional World Model","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15157","citing_title":"Hand-in-the-Loop: Improving VLA Policies for Dexterous Manipulation via Seamless Hand-Arm Intervention","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03037","citing_title":"ARM: Advantage Reward Modeling for Long-Horizon Manipulation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2511.14759","citing_title":"$\\pi^{*}_{0.6}$: a VLA That Learns From Experience","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10821","citing_title":"UniSteer: Unified Noise Steering for Efficient Human-Guided VLA Adaptation","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07335","citing_title":"TAMEn: Tactile-Aware Manipulation Engine for Closed-Loop Data Collection in Contact-Rich Tasks","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD","json":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD.json","graph_json":"https://pith.science/api/pith-number/2SZANLNDCKHQJ5JQCZJTLIDITD/graph.json","events_json":"https://pith.science/api/pith-number/2SZANLNDCKHQJ5JQCZJTLIDITD/events.json","paper":"https://pith.science/paper/2SZANLND"},"agent_actions":{"view_html":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD","download_json":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD.json","view_paper":"https://pith.science/paper/2SZANLND","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.07953&json=true","fetch_graph":"https://pith.science/api/pith-number/2SZANLNDCKHQJ5JQCZJTLIDITD/graph.json","fetch_events":"https://pith.science/api/pith-number/2SZANLNDCKHQJ5JQCZJTLIDITD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD/action/storage_attestation","attest_author":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD/action/author_attestation","sign_citation":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD/action/citation_signature","submit_replication":"https://pith.science/pith/2SZANLNDCKHQJ5JQCZJTLIDITD/action/replication_record"}},"created_at":"2026-07-05T12:07:42.957758+00:00","updated_at":"2026-07-05T12:07:42.957758+00:00"}