{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:BR2KXGU2K4Y4JYBPV6ZRQDWKEF","short_pith_number":"pith:BR2KXGU2","schema_version":"1.0","canonical_sha256":"0c74ab9a9a5731c4e02fafb3180eca214a76dde516df426756320cfab9d6a777","source":{"kind":"arxiv","id":"2102.02872","version":2},"attestation_state":"computed","paper":{"title":"Feedback in Imitation Learning: The Three Regimes of Covariate Shift","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.RO","stat.ML"],"primary_cat":"cs.LG","authors_text":"Arun Venkatraman, Brian Ziebart, J. Andrew Bagnell, Jonathan Spencer, Sanjiban Choudhury","submitted_at":"2021-02-04T20:18:56Z","abstract_excerpt":"Imitation learning practitioners have often noted that conditioning policies on previous actions leads to a dramatic divergence between \"held out\" error and performance of the learner in situ. Interactive approaches can provably address this divergence but require repeated querying of a demonstrator. Recent work identifies this divergence as stemming from a \"causal confound\" in predicting the current action, and seek to ablate causal aspects of current state using tools from causal inference. In this work, we argue instead that this divergence is simply another manifestation of covariate shift"},"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":"2102.02872","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-02-04T20:18:56Z","cross_cats_sorted":["cs.RO","stat.ML"],"title_canon_sha256":"c7e0395ce9e6e5675dbd2597c86e8a48b8031114c998a76e65823c3f16d24ac4","abstract_canon_sha256":"69048dbab7ae9cdbdd41df6f3379f767b0b50d1b52be64234135449b3f98e893"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:14:33.399727Z","signature_b64":"MPkG37eOJ0Wfz6+ViUq9Jzv3DjSdopxroFj3DxW8r+undtM+gQyw5w0GI5ZnqT2xY5wBgZ/F7nLza7XSENoTCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0c74ab9a9a5731c4e02fafb3180eca214a76dde516df426756320cfab9d6a777","last_reissued_at":"2026-07-05T02:14:33.399276Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:14:33.399276Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Feedback in Imitation Learning: The Three Regimes of Covariate Shift","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.RO","stat.ML"],"primary_cat":"cs.LG","authors_text":"Arun Venkatraman, Brian Ziebart, J. Andrew Bagnell, Jonathan Spencer, Sanjiban Choudhury","submitted_at":"2021-02-04T20:18:56Z","abstract_excerpt":"Imitation learning practitioners have often noted that conditioning policies on previous actions leads to a dramatic divergence between \"held out\" error and performance of the learner in situ. Interactive approaches can provably address this divergence but require repeated querying of a demonstrator. Recent work identifies this divergence as stemming from a \"causal confound\" in predicting the current action, and seek to ablate causal aspects of current state using tools from causal inference. In this work, we argue instead that this divergence is simply another manifestation of covariate shift"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.02872","kind":"arxiv","version":2},"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/2102.02872/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":"2102.02872","created_at":"2026-07-05T02:14:33.399344+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.02872v2","created_at":"2026-07-05T02:14:33.399344+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.02872","created_at":"2026-07-05T02:14:33.399344+00:00"},{"alias_kind":"pith_short_12","alias_value":"BR2KXGU2K4Y4","created_at":"2026-07-05T02:14:33.399344+00:00"},{"alias_kind":"pith_short_16","alias_value":"BR2KXGU2K4Y4JYBP","created_at":"2026-07-05T02:14:33.399344+00:00"},{"alias_kind":"pith_short_8","alias_value":"BR2KXGU2","created_at":"2026-07-05T02:14:33.399344+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25136","citing_title":"Memory Retrieval in Visuomotor Policies for Long-Horizon Robot Control","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09758","citing_title":"Difference-Aware Retrieval Policies for Imitation Learning","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2512.05812","citing_title":"Toward Efficient and Robust Behavior Models for Multi-Agent Driving Simulation","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14484","citing_title":"Behavior Cloning Under PD Control: A Finite-Horizon Theory of Gain-Dependent Error Amplification","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09153","citing_title":"Beyond Self-Play: Hierarchical Reasoning for Continuous Motion in Closed-Loop Traffic Simulation","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14484","citing_title":"Behavior Cloning Under PD Control: A Finite-Horizon Theory of Gain-Dependent Error Amplification","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF","json":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF.json","graph_json":"https://pith.science/api/pith-number/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/graph.json","events_json":"https://pith.science/api/pith-number/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/events.json","paper":"https://pith.science/paper/BR2KXGU2"},"agent_actions":{"view_html":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF","download_json":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF.json","view_paper":"https://pith.science/paper/BR2KXGU2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.02872&json=true","fetch_graph":"https://pith.science/api/pith-number/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/graph.json","fetch_events":"https://pith.science/api/pith-number/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/action/storage_attestation","attest_author":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/action/author_attestation","sign_citation":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/action/citation_signature","submit_replication":"https://pith.science/pith/BR2KXGU2K4Y4JYBPV6ZRQDWKEF/action/replication_record"}},"created_at":"2026-07-05T02:14:33.399344+00:00","updated_at":"2026-07-05T02:14:33.399344+00:00"}