{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YWJKGJLSZIW3LHMX5KNZG5PQGE","short_pith_number":"pith:YWJKGJLS","schema_version":"1.0","canonical_sha256":"c592a32572ca2db59d97ea9b9375f0310c253c051d642c15f9915328074e014d","source":{"kind":"arxiv","id":"2505.09561","version":2},"attestation_state":"computed","paper":{"title":"Learning Long-Context Diffusion Policies via Past-Token Prediction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.RO","authors_text":"Andy Tang, Chelsea Finn, Marcel Torne, Yuejiang Liu","submitted_at":"2025-05-14T17:00:47Z","abstract_excerpt":"Reasoning over long sequences of observations and actions is essential for many robotic tasks. Yet, learning effective long-context policies from demonstrations remains challenging. As context length increases, training becomes increasingly expensive due to rising memory demands, and policy performance often degrades as a result of spurious correlations. Recent methods typically sidestep these issues by truncating context length, discarding historical information that may be critical for subsequent decisions. In this paper, we propose an alternative approach that explicitly regularizes the ret"},"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":"2505.09561","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-05-14T17:00:47Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"4c670ad85178a573da019e6e5d730dd670b09cbc54901abfe9e49c73f9cf43f0","abstract_canon_sha256":"c805613d050cfeca40c95519bcf628f6e4a14092da0951f65b742fce04e3856c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:05:44.477831Z","signature_b64":"3wZm3SLwmTS8jfvTMywBf2CU3gF37T0qJ8BnyMP5nnQ5KbtE1/k+YxtaXVssTpPprjl/y6xWPnuD0AMgnHicCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c592a32572ca2db59d97ea9b9375f0310c253c051d642c15f9915328074e014d","last_reissued_at":"2026-07-05T11:05:44.477327Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:05:44.477327Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Long-Context Diffusion Policies via Past-Token Prediction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.RO","authors_text":"Andy Tang, Chelsea Finn, Marcel Torne, Yuejiang Liu","submitted_at":"2025-05-14T17:00:47Z","abstract_excerpt":"Reasoning over long sequences of observations and actions is essential for many robotic tasks. Yet, learning effective long-context policies from demonstrations remains challenging. As context length increases, training becomes increasingly expensive due to rising memory demands, and policy performance often degrades as a result of spurious correlations. Recent methods typically sidestep these issues by truncating context length, discarding historical information that may be critical for subsequent decisions. In this paper, we propose an alternative approach that explicitly regularizes the ret"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.09561","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/2505.09561/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":"2505.09561","created_at":"2026-07-05T11:05:44.477390+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.09561v2","created_at":"2026-07-05T11:05:44.477390+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.09561","created_at":"2026-07-05T11:05:44.477390+00:00"},{"alias_kind":"pith_short_12","alias_value":"YWJKGJLSZIW3","created_at":"2026-07-05T11:05:44.477390+00:00"},{"alias_kind":"pith_short_16","alias_value":"YWJKGJLSZIW3LHMX","created_at":"2026-07-05T11:05:44.477390+00:00"},{"alias_kind":"pith_short_8","alias_value":"YWJKGJLS","created_at":"2026-07-05T11:05:44.477390+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21188","citing_title":"Remember what you did?: Learning Behavioral Memories for Partially Observable Object Manipulation","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20092","citing_title":"EventVLA: Event-Driven Visual Evidence Memory for Long-Horizon Vision-Language-Action Policies","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01072","citing_title":"Expanding Spatial and Temporal Context for Robotic Imitation Learning With Scene Graphs","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20092","citing_title":"EventVLA: Event-Driven Visual Evidence Memory for Long-Horizon Vision-Language-Action Policies","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14598","citing_title":"DSSP: Diffusion State Space Policy with Full-History Encoding","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2603.04639","citing_title":"RoboMME: Benchmarking and Understanding Memory for Robotic Generalist Policies","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03269","citing_title":"RLDX-1 Technical Report","ref_index":100,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00814","citing_title":"Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18933","citing_title":"Gated Memory Policy: In-Context Memorization and Adaptation","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00814","citing_title":"Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03269","citing_title":"RLDX-1 Technical Report","ref_index":100,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE","json":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE.json","graph_json":"https://pith.science/api/pith-number/YWJKGJLSZIW3LHMX5KNZG5PQGE/graph.json","events_json":"https://pith.science/api/pith-number/YWJKGJLSZIW3LHMX5KNZG5PQGE/events.json","paper":"https://pith.science/paper/YWJKGJLS"},"agent_actions":{"view_html":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE","download_json":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE.json","view_paper":"https://pith.science/paper/YWJKGJLS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.09561&json=true","fetch_graph":"https://pith.science/api/pith-number/YWJKGJLSZIW3LHMX5KNZG5PQGE/graph.json","fetch_events":"https://pith.science/api/pith-number/YWJKGJLSZIW3LHMX5KNZG5PQGE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE/action/storage_attestation","attest_author":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE/action/author_attestation","sign_citation":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE/action/citation_signature","submit_replication":"https://pith.science/pith/YWJKGJLSZIW3LHMX5KNZG5PQGE/action/replication_record"}},"created_at":"2026-07-05T11:05:44.477390+00:00","updated_at":"2026-07-05T11:05:44.477390+00:00"}