{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XKH34AEJVDQY4TOSWMPDZ7RGLE","short_pith_number":"pith:XKH34AEJ","schema_version":"1.0","canonical_sha256":"ba8fbe0089a8e18e4dd2b31e3cfe265900b7f4054181f34bbdc9ddfac1ee4042","source":{"kind":"arxiv","id":"2410.21035","version":2},"attestation_state":"computed","paper":{"title":"Beyond Autoregression: Fast LLMs via Self-Distillation Through Time","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Caglar Gulcehre, Justin Deschenaux","submitted_at":"2024-10-28T13:56:30Z","abstract_excerpt":"Autoregressive (AR) Large Language Models (LLMs) have demonstrated significant success across numerous tasks. However, the AR modeling paradigm presents certain limitations; for instance, contemporary autoregressive LLMs are trained to generate one token at a time, which can result in noticeable latency. Recent advances have indicated that search and repeated sampling can enhance performance in various applications, such as theorem proving, code generation, and alignment, by utilizing greater computational resources during inference. In this study, we demonstrate that diffusion language models"},"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":"2410.21035","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-28T13:56:30Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"6adaf4a4b95369a37b5b09ccaadfde79c06ac8753522671ebd9dc37d7dc40f15","abstract_canon_sha256":"0cae885210f04a8bdcbc13ff85184c51aad736152cb8fdc37b4899d2c5a35669"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:10:45.451459Z","signature_b64":"x+sObk/m84Or5PPhWwYxlQfo+GdPT9UsBv2euG8I0A0+163wuK1cIg7ZXk9YUHPSOINSp3oB6cdAlFe/DGZTAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba8fbe0089a8e18e4dd2b31e3cfe265900b7f4054181f34bbdc9ddfac1ee4042","last_reissued_at":"2026-07-05T10:10:45.451015Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:10:45.451015Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Autoregression: Fast LLMs via Self-Distillation Through Time","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Caglar Gulcehre, Justin Deschenaux","submitted_at":"2024-10-28T13:56:30Z","abstract_excerpt":"Autoregressive (AR) Large Language Models (LLMs) have demonstrated significant success across numerous tasks. However, the AR modeling paradigm presents certain limitations; for instance, contemporary autoregressive LLMs are trained to generate one token at a time, which can result in noticeable latency. Recent advances have indicated that search and repeated sampling can enhance performance in various applications, such as theorem proving, code generation, and alignment, by utilizing greater computational resources during inference. In this study, we demonstrate that diffusion language models"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.21035","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/2410.21035/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":"2410.21035","created_at":"2026-07-05T10:10:45.451071+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.21035v2","created_at":"2026-07-05T10:10:45.451071+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.21035","created_at":"2026-07-05T10:10:45.451071+00:00"},{"alias_kind":"pith_short_12","alias_value":"XKH34AEJVDQY","created_at":"2026-07-05T10:10:45.451071+00:00"},{"alias_kind":"pith_short_16","alias_value":"XKH34AEJVDQY4TOS","created_at":"2026-07-05T10:10:45.451071+00:00"},{"alias_kind":"pith_short_8","alias_value":"XKH34AEJ","created_at":"2026-07-05T10:10:45.451071+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.24773","citing_title":"Posterior Refinement: Fast Language Generation via Any-Order Flow Maps","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00714","citing_title":"Self-conditioned Flow Map Language Models via Fixed-point Flows","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27617","citing_title":"Masked Language Flow Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31215","citing_title":"Fixed-Point Masked Generative Modeling","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2602.16813","citing_title":"Flow Map Language Models: One-step Language Modeling via Continuous Denoising","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21484","citing_title":"One-Step Distillation of Discrete Diffusion Image Generators via Fixed-Point Iteration","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2602.16813","citing_title":"Flow Map Language Models: One-step Language Modeling via Continuous Denoising","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07193","citing_title":"Coupling Models for One-Step Discrete Generation","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2502.09992","citing_title":"Large Language Diffusion Models","ref_index":88,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE","json":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE.json","graph_json":"https://pith.science/api/pith-number/XKH34AEJVDQY4TOSWMPDZ7RGLE/graph.json","events_json":"https://pith.science/api/pith-number/XKH34AEJVDQY4TOSWMPDZ7RGLE/events.json","paper":"https://pith.science/paper/XKH34AEJ"},"agent_actions":{"view_html":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE","download_json":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE.json","view_paper":"https://pith.science/paper/XKH34AEJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.21035&json=true","fetch_graph":"https://pith.science/api/pith-number/XKH34AEJVDQY4TOSWMPDZ7RGLE/graph.json","fetch_events":"https://pith.science/api/pith-number/XKH34AEJVDQY4TOSWMPDZ7RGLE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE/action/storage_attestation","attest_author":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE/action/author_attestation","sign_citation":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE/action/citation_signature","submit_replication":"https://pith.science/pith/XKH34AEJVDQY4TOSWMPDZ7RGLE/action/replication_record"}},"created_at":"2026-07-05T10:10:45.451071+00:00","updated_at":"2026-07-05T10:10:45.451071+00:00"}