{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:6ESMROJMVG33HH5S5ZB2PUZ57Y","short_pith_number":"pith:6ESMROJM","schema_version":"1.0","canonical_sha256":"f124c8b92ca9b7b39fb2ee43a7d33dfe037397d91b7fca0e6439a7e26feb536c","source":{"kind":"arxiv","id":"2606.08015","version":1},"attestation_state":"computed","paper":{"title":"Q-VGM: Q-Guided Value-Gradient Matching for Flow-Matching VLA Policies","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Jiayu Sun, Minqian Wang, Xingjian Mao, Yao Mu, Ziqian Wang","submitted_at":"2026-06-06T07:10:25Z","abstract_excerpt":"We propose Q-Guided Value-Gradient Matching (Q-VGM), an off-policy reinforcement learning (RL) method that tackles a long-standing challenge in fine-tuning flow-matching vision-language-action (VLA) policies: efficiently improving an expressive flow-matching action expert with respect to a learned Q-function. Effective improvement must exploit the first-order (gradient) information of the critic, but this is difficult for flow policies, because directly back-propagating the value through their multi-step denoising process is numerically unstable at VLA scale, while the tractable action likelih"},"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":"2606.08015","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-06-06T07:10:25Z","cross_cats_sorted":[],"title_canon_sha256":"20ce601ffb35df96f1f6dd2098a492386e49d1ae29a0ec37d7838b92530918df","abstract_canon_sha256":"d3bc3aa72d96eb1d0720a95dd1db140cd5f83c1ae0268b2151a5da93ce19ed35"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:05:22.823286Z","signature_b64":"1v9thzynZrjCuYsVR14QGW9T1A+hXr31EzWzabOFWAeHzWWNs7ApPPP59a5uW7sG991wDWdlIrMLN5UL2l+YAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f124c8b92ca9b7b39fb2ee43a7d33dfe037397d91b7fca0e6439a7e26feb536c","last_reissued_at":"2026-06-09T01:05:22.822521Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:05:22.822521Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Q-VGM: Q-Guided Value-Gradient Matching for Flow-Matching VLA Policies","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Jiayu Sun, Minqian Wang, Xingjian Mao, Yao Mu, Ziqian Wang","submitted_at":"2026-06-06T07:10:25Z","abstract_excerpt":"We propose Q-Guided Value-Gradient Matching (Q-VGM), an off-policy reinforcement learning (RL) method that tackles a long-standing challenge in fine-tuning flow-matching vision-language-action (VLA) policies: efficiently improving an expressive flow-matching action expert with respect to a learned Q-function. Effective improvement must exploit the first-order (gradient) information of the critic, but this is difficult for flow policies, because directly back-propagating the value through their multi-step denoising process is numerically unstable at VLA scale, while the tractable action likelih"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.08015","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/2606.08015/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":"2606.08015","created_at":"2026-06-09T01:05:22.822630+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.08015v1","created_at":"2026-06-09T01:05:22.822630+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.08015","created_at":"2026-06-09T01:05:22.822630+00:00"},{"alias_kind":"pith_short_12","alias_value":"6ESMROJMVG33","created_at":"2026-06-09T01:05:22.822630+00:00"},{"alias_kind":"pith_short_16","alias_value":"6ESMROJMVG33HH5S","created_at":"2026-06-09T01:05:22.822630+00:00"},{"alias_kind":"pith_short_8","alias_value":"6ESMROJM","created_at":"2026-06-09T01:05:22.822630+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/6ESMROJMVG33HH5S5ZB2PUZ57Y","json":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y.json","graph_json":"https://pith.science/api/pith-number/6ESMROJMVG33HH5S5ZB2PUZ57Y/graph.json","events_json":"https://pith.science/api/pith-number/6ESMROJMVG33HH5S5ZB2PUZ57Y/events.json","paper":"https://pith.science/paper/6ESMROJM"},"agent_actions":{"view_html":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y","download_json":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y.json","view_paper":"https://pith.science/paper/6ESMROJM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.08015&json=true","fetch_graph":"https://pith.science/api/pith-number/6ESMROJMVG33HH5S5ZB2PUZ57Y/graph.json","fetch_events":"https://pith.science/api/pith-number/6ESMROJMVG33HH5S5ZB2PUZ57Y/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y/action/storage_attestation","attest_author":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y/action/author_attestation","sign_citation":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y/action/citation_signature","submit_replication":"https://pith.science/pith/6ESMROJMVG33HH5S5ZB2PUZ57Y/action/replication_record"}},"created_at":"2026-06-09T01:05:22.822630+00:00","updated_at":"2026-06-09T01:05:22.822630+00:00"}