{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7QR5RDTOGUBLRGFRGGW4CFKIKZ","short_pith_number":"pith:7QR5RDTO","schema_version":"1.0","canonical_sha256":"fc23d88e6e3502b898b131adc11548567bdf515812b5bbe1af2b8108e9c94cf7","source":{"kind":"arxiv","id":"2605.23605","version":1},"attestation_state":"computed","paper":{"title":"DiLaDiff: Distilled Latent-Augmented Diffusion for Language Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Ante Juki\\'c, Arash Vahdat, Jean-Marie Lemercier, Karsten Kreis, Morteza Mardani, Tomas Geffner","submitted_at":"2026-05-22T13:15:59Z","abstract_excerpt":"Diffusion language models intrinsically fail to capture correlations between decoded tokens, which leads to a harsh trade-off between sampling quality and throughput. To solve this issue, we propose DiLaDiff, a variant of masked diffusion language models with three components: (1) a continuous latent space with semantic capabilities, learned by an auto-encoder fine-tuned from an existing masked diffusion language model; (2) a latent diffusion model learning the prior over the encoder distribution; (3) a consistency model distilling the learned prior into a few-step latent generative model. We "},"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":"2605.23605","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-22T13:15:59Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"0ab650c6f7ad0e8a6b4fdbc5564021e3e4f38603a125faddf03d20d2b32c3eae","abstract_canon_sha256":"71bf226813c5d59c369294f9a93b14b22690d16cb9977c50b3ac2201147239c0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:02:21.815091Z","signature_b64":"HYiqsdI7hUr02d1EJDfmuorIFqxFqyqpQXFIU0H9p/OBrwrC/ky3zflWzwqG82iflex3wF+RU2hREhhabf2+CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fc23d88e6e3502b898b131adc11548567bdf515812b5bbe1af2b8108e9c94cf7","last_reissued_at":"2026-05-25T02:02:21.814282Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:02:21.814282Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DiLaDiff: Distilled Latent-Augmented Diffusion for Language Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Ante Juki\\'c, Arash Vahdat, Jean-Marie Lemercier, Karsten Kreis, Morteza Mardani, Tomas Geffner","submitted_at":"2026-05-22T13:15:59Z","abstract_excerpt":"Diffusion language models intrinsically fail to capture correlations between decoded tokens, which leads to a harsh trade-off between sampling quality and throughput. To solve this issue, we propose DiLaDiff, a variant of masked diffusion language models with three components: (1) a continuous latent space with semantic capabilities, learned by an auto-encoder fine-tuned from an existing masked diffusion language model; (2) a latent diffusion model learning the prior over the encoder distribution; (3) a consistency model distilling the learned prior into a few-step latent generative model. We "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.23605","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/2605.23605/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":"2605.23605","created_at":"2026-05-25T02:02:21.814426+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.23605v1","created_at":"2026-05-25T02:02:21.814426+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.23605","created_at":"2026-05-25T02:02:21.814426+00:00"},{"alias_kind":"pith_short_12","alias_value":"7QR5RDTOGUBL","created_at":"2026-05-25T02:02:21.814426+00:00"},{"alias_kind":"pith_short_16","alias_value":"7QR5RDTOGUBLRGFR","created_at":"2026-05-25T02:02:21.814426+00:00"},{"alias_kind":"pith_short_8","alias_value":"7QR5RDTO","created_at":"2026-05-25T02:02:21.814426+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/7QR5RDTOGUBLRGFRGGW4CFKIKZ","json":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ.json","graph_json":"https://pith.science/api/pith-number/7QR5RDTOGUBLRGFRGGW4CFKIKZ/graph.json","events_json":"https://pith.science/api/pith-number/7QR5RDTOGUBLRGFRGGW4CFKIKZ/events.json","paper":"https://pith.science/paper/7QR5RDTO"},"agent_actions":{"view_html":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ","download_json":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ.json","view_paper":"https://pith.science/paper/7QR5RDTO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.23605&json=true","fetch_graph":"https://pith.science/api/pith-number/7QR5RDTOGUBLRGFRGGW4CFKIKZ/graph.json","fetch_events":"https://pith.science/api/pith-number/7QR5RDTOGUBLRGFRGGW4CFKIKZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ/action/storage_attestation","attest_author":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ/action/author_attestation","sign_citation":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ/action/citation_signature","submit_replication":"https://pith.science/pith/7QR5RDTOGUBLRGFRGGW4CFKIKZ/action/replication_record"}},"created_at":"2026-05-25T02:02:21.814426+00:00","updated_at":"2026-05-25T02:02:21.814426+00:00"}