{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CE4MWMQ44UMEFFDA6UE2CWEXEL","short_pith_number":"pith:CE4MWMQ4","schema_version":"1.0","canonical_sha256":"1138cb321ce518429460f509a1589722e7fde1c321ed3ee10963dfc01dae5231","source":{"kind":"arxiv","id":"2601.08136","version":2},"attestation_state":"computed","paper":{"title":"Reverse Flow Matching: A Unified Framework for Online Reinforcement Learning with Diffusion and Flow Policies","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SY","eess.SY"],"primary_cat":"cs.LG","authors_text":"Navid Azizan, Sunbochen Tang, Zeyang Li","submitted_at":"2026-01-13T01:58:24Z","abstract_excerpt":"Diffusion and flow policies are gaining prominence in online reinforcement learning (RL) due to their expressive power, yet training them efficiently remains a critical challenge. A fundamental difficulty that distinguishes online RL from standard generative modeling is the lack of direct samples from the target Boltzmann distribution defined by the Q-function. To address this, two seemingly distinct families of methods have been proposed for diffusion policies: a noise-expectation family, which uses a weighted average of noise as the training target, and a gradient-expectation family, which e"},"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":"2601.08136","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-01-13T01:58:24Z","cross_cats_sorted":["cs.SY","eess.SY"],"title_canon_sha256":"958f739560d62c66dbd3ca63e2395aa27ca0c550e1d5776132e3c5ed9f631ae0","abstract_canon_sha256":"c93e7569265b9c169fcb060471b8e111c4362a399ae439cd23b2d943ba91b8a4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-11T01:09:26.586624Z","signature_b64":"ob6Luc2pqdiL+4qVrDg0RccKX9ZihEAsUniru3QVoAegNaltbdbWQcNC2vFtLF7PV4a+P8N/tUrguvFfmLVPCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1138cb321ce518429460f509a1589722e7fde1c321ed3ee10963dfc01dae5231","last_reissued_at":"2026-06-11T01:09:26.586045Z","signature_status":"signed_v1","first_computed_at":"2026-06-11T01:09:26.586045Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reverse Flow Matching: A Unified Framework for Online Reinforcement Learning with Diffusion and Flow Policies","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SY","eess.SY"],"primary_cat":"cs.LG","authors_text":"Navid Azizan, Sunbochen Tang, Zeyang Li","submitted_at":"2026-01-13T01:58:24Z","abstract_excerpt":"Diffusion and flow policies are gaining prominence in online reinforcement learning (RL) due to their expressive power, yet training them efficiently remains a critical challenge. A fundamental difficulty that distinguishes online RL from standard generative modeling is the lack of direct samples from the target Boltzmann distribution defined by the Q-function. To address this, two seemingly distinct families of methods have been proposed for diffusion policies: a noise-expectation family, which uses a weighted average of noise as the training target, and a gradient-expectation family, which e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.08136","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/2601.08136/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":"2601.08136","created_at":"2026-06-11T01:09:26.586113+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.08136v2","created_at":"2026-06-11T01:09:26.586113+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.08136","created_at":"2026-06-11T01:09:26.586113+00:00"},{"alias_kind":"pith_short_12","alias_value":"CE4MWMQ44UME","created_at":"2026-06-11T01:09:26.586113+00:00"},{"alias_kind":"pith_short_16","alias_value":"CE4MWMQ44UMEFFDA","created_at":"2026-06-11T01:09:26.586113+00:00"},{"alias_kind":"pith_short_8","alias_value":"CE4MWMQ4","created_at":"2026-06-11T01:09:26.586113+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL","json":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL.json","graph_json":"https://pith.science/api/pith-number/CE4MWMQ44UMEFFDA6UE2CWEXEL/graph.json","events_json":"https://pith.science/api/pith-number/CE4MWMQ44UMEFFDA6UE2CWEXEL/events.json","paper":"https://pith.science/paper/CE4MWMQ4"},"agent_actions":{"view_html":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL","download_json":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL.json","view_paper":"https://pith.science/paper/CE4MWMQ4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.08136&json=true","fetch_graph":"https://pith.science/api/pith-number/CE4MWMQ44UMEFFDA6UE2CWEXEL/graph.json","fetch_events":"https://pith.science/api/pith-number/CE4MWMQ44UMEFFDA6UE2CWEXEL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL/action/storage_attestation","attest_author":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL/action/author_attestation","sign_citation":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL/action/citation_signature","submit_replication":"https://pith.science/pith/CE4MWMQ44UMEFFDA6UE2CWEXEL/action/replication_record"}},"created_at":"2026-06-11T01:09:26.586113+00:00","updated_at":"2026-06-11T01:09:26.586113+00:00"}