{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5VNZEUMT3SA7NNCW2HZH2VMEVA","short_pith_number":"pith:5VNZEUMT","schema_version":"1.0","canonical_sha256":"ed5b925193dc81f6b456d1f27d5584a80a48497891b09e8f91d6f2e1b0ccd14a","source":{"kind":"arxiv","id":"2606.27617","version":1},"attestation_state":"computed","paper":{"title":"Masked Language Flow Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Iskander Azangulov, Kianoosh Ashouritaklimi, Leo Zhang, Patrick Rebeschini, Simon Vary","submitted_at":"2026-06-26T00:16:40Z","abstract_excerpt":"Masked Diffusion Models (MDMs) promise fast, parallel language generation, but their reverse transition factorises across token positions -- an approximation that breaks down in the few-step sampling regime where parallel generation ought to provide the greatest efficiency gains. Flow Language Models (FLMs) sidestep this limitation by learning a continuous flow that transports noise toward clean sequences represented in Euclidean space, inducing a flow map that can be distilled for single-step generation. However, this makes complex tasks requiring multi-step reasoning problematic for FLMs, as"},"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":"2606.27617","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-26T00:16:40Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4536675d9432dd68c4151f0332350f1acdbd523c0b6ed402abe4e50722e934de","abstract_canon_sha256":"8038844a3d327663a3004753e3951f59118d0e509923402d2277bb58d0f25d4c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-29T00:14:12.034792Z","signature_b64":"6EXLjsq43Y3CoDaa/UWHqrFAa7gijZyLEYC/YaFy5BnFSvseGud3gROmd0gKCBpqqd+8quJu/W2hVvTzISfZBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ed5b925193dc81f6b456d1f27d5584a80a48497891b09e8f91d6f2e1b0ccd14a","last_reissued_at":"2026-06-29T00:14:12.034350Z","signature_status":"signed_v1","first_computed_at":"2026-06-29T00:14:12.034350Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Masked Language Flow Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Iskander Azangulov, Kianoosh Ashouritaklimi, Leo Zhang, Patrick Rebeschini, Simon Vary","submitted_at":"2026-06-26T00:16:40Z","abstract_excerpt":"Masked Diffusion Models (MDMs) promise fast, parallel language generation, but their reverse transition factorises across token positions -- an approximation that breaks down in the few-step sampling regime where parallel generation ought to provide the greatest efficiency gains. Flow Language Models (FLMs) sidestep this limitation by learning a continuous flow that transports noise toward clean sequences represented in Euclidean space, inducing a flow map that can be distilled for single-step generation. However, this makes complex tasks requiring multi-step reasoning problematic for FLMs, as"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27617","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/2606.27617/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":"2606.27617","created_at":"2026-06-29T00:14:12.034419+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.27617v1","created_at":"2026-06-29T00:14:12.034419+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27617","created_at":"2026-06-29T00:14:12.034419+00:00"},{"alias_kind":"pith_short_12","alias_value":"5VNZEUMT3SA7","created_at":"2026-06-29T00:14:12.034419+00:00"},{"alias_kind":"pith_short_16","alias_value":"5VNZEUMT3SA7NNCW","created_at":"2026-06-29T00:14:12.034419+00:00"},{"alias_kind":"pith_short_8","alias_value":"5VNZEUMT","created_at":"2026-06-29T00:14:12.034419+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/5VNZEUMT3SA7NNCW2HZH2VMEVA","json":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA.json","graph_json":"https://pith.science/api/pith-number/5VNZEUMT3SA7NNCW2HZH2VMEVA/graph.json","events_json":"https://pith.science/api/pith-number/5VNZEUMT3SA7NNCW2HZH2VMEVA/events.json","paper":"https://pith.science/paper/5VNZEUMT"},"agent_actions":{"view_html":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA","download_json":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA.json","view_paper":"https://pith.science/paper/5VNZEUMT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.27617&json=true","fetch_graph":"https://pith.science/api/pith-number/5VNZEUMT3SA7NNCW2HZH2VMEVA/graph.json","fetch_events":"https://pith.science/api/pith-number/5VNZEUMT3SA7NNCW2HZH2VMEVA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA/action/storage_attestation","attest_author":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA/action/author_attestation","sign_citation":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA/action/citation_signature","submit_replication":"https://pith.science/pith/5VNZEUMT3SA7NNCW2HZH2VMEVA/action/replication_record"}},"created_at":"2026-06-29T00:14:12.034419+00:00","updated_at":"2026-06-29T00:14:12.034419+00:00"}