{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:6YUAHKKEHRJYVHWKATWCGLBR77","short_pith_number":"pith:6YUAHKKE","schema_version":"1.0","canonical_sha256":"f62803a9443c538a9eca04ec232c31ffe75febaaf0774c4eb514e9ce843a00c3","source":{"kind":"arxiv","id":"2602.19066","version":2},"attestation_state":"computed","paper":{"title":"IDLM: Inverse-distilled Diffusion Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexander Korotin, David Li, Dmitry Abulkhanov, Eric Moulines, Ivan Oseledets, Maxim Panov, Nikita Gushchin","submitted_at":"2026-02-22T06:47:04Z","abstract_excerpt":"Diffusion Language Models (DLMs) have recently achieved strong results in text generation. However, their multi-step sampling leads to slow inference, limiting practical use. To address this, we extend Inverse Distillation, a technique originally developed to accelerate continuous diffusion models, to the discrete setting. Nonetheless, this extension introduces both theoretical and practical challenges. From a theoretical perspective, the inverse distillation objective lacks uniqueness guarantees, which may lead to suboptimal solutions. From a practical standpoint, backpropagation in the discr"},"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":"2602.19066","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-22T06:47:04Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9559e8a22c92355639515afd21e899d09592444eb782964303d0ca0e2eef061c","abstract_canon_sha256":"54ba55503b739f04a0534da43c3e28e61ac058888668b32f2d1981cc1b065f40"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:43.979783Z","signature_b64":"CZxtEahbeT5TJq+2el6MP5zNSywkY0MBd5mRqpp5cr/rpPYdEoMLp6GzYStDvY4MwAuOAAX2wghqa4TlvrPrDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f62803a9443c538a9eca04ec232c31ffe75febaaf0774c4eb514e9ce843a00c3","last_reissued_at":"2026-06-02T01:03:43.979171Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:43.979171Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"IDLM: Inverse-distilled Diffusion Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexander Korotin, David Li, Dmitry Abulkhanov, Eric Moulines, Ivan Oseledets, Maxim Panov, Nikita Gushchin","submitted_at":"2026-02-22T06:47:04Z","abstract_excerpt":"Diffusion Language Models (DLMs) have recently achieved strong results in text generation. However, their multi-step sampling leads to slow inference, limiting practical use. To address this, we extend Inverse Distillation, a technique originally developed to accelerate continuous diffusion models, to the discrete setting. Nonetheless, this extension introduces both theoretical and practical challenges. From a theoretical perspective, the inverse distillation objective lacks uniqueness guarantees, which may lead to suboptimal solutions. From a practical standpoint, backpropagation in the discr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.19066","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/2602.19066/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":"2602.19066","created_at":"2026-06-02T01:03:43.979229+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.19066v2","created_at":"2026-06-02T01:03:43.979229+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.19066","created_at":"2026-06-02T01:03:43.979229+00:00"},{"alias_kind":"pith_short_12","alias_value":"6YUAHKKEHRJY","created_at":"2026-06-02T01:03:43.979229+00:00"},{"alias_kind":"pith_short_16","alias_value":"6YUAHKKEHRJYVHWK","created_at":"2026-06-02T01:03:43.979229+00:00"},{"alias_kind":"pith_short_8","alias_value":"6YUAHKKE","created_at":"2026-06-02T01:03:43.979229+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.31215","citing_title":"Fixed-Point Masked Generative Modeling","ref_index":43,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77","json":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77.json","graph_json":"https://pith.science/api/pith-number/6YUAHKKEHRJYVHWKATWCGLBR77/graph.json","events_json":"https://pith.science/api/pith-number/6YUAHKKEHRJYVHWKATWCGLBR77/events.json","paper":"https://pith.science/paper/6YUAHKKE"},"agent_actions":{"view_html":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77","download_json":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77.json","view_paper":"https://pith.science/paper/6YUAHKKE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.19066&json=true","fetch_graph":"https://pith.science/api/pith-number/6YUAHKKEHRJYVHWKATWCGLBR77/graph.json","fetch_events":"https://pith.science/api/pith-number/6YUAHKKEHRJYVHWKATWCGLBR77/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77/action/storage_attestation","attest_author":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77/action/author_attestation","sign_citation":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77/action/citation_signature","submit_replication":"https://pith.science/pith/6YUAHKKEHRJYVHWKATWCGLBR77/action/replication_record"}},"created_at":"2026-06-02T01:03:43.979229+00:00","updated_at":"2026-06-02T01:03:43.979229+00:00"}