{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:IXWHSV5RPBRRGG2KA6IWQXVQA5","short_pith_number":"pith:IXWHSV5R","schema_version":"1.0","canonical_sha256":"45ec7957b17863131b4a0791685eb0077daec241541b2427f16c9269b93b6a59","source":{"kind":"arxiv","id":"2501.15781","version":2},"attestation_state":"computed","paper":{"title":"Large Language Models to Diffusion Finetuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Edoardo Cetin, Tianyu Zhao, Yujin Tang","submitted_at":"2025-01-27T04:59:29Z","abstract_excerpt":"We propose a new finetuning method to provide pre-trained large language models (LMs) the ability to scale test-time compute through the diffusion framework. By increasing the number of diffusion steps, we show our finetuned models achieve monotonically increasing accuracy, directly translating to improved performance across downstream tasks. Furthermore, our finetuned models can expertly answer questions on specific topics by integrating powerful guidance techniques, and autonomously determine the compute required for a given problem by leveraging adaptive ODE solvers. Our method is universal"},"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":"2501.15781","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-27T04:59:29Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"1d4bfb9670940c8a8f40917dafb367bd05c3d8e000447cfe7f4a951780b8530d","abstract_canon_sha256":"abf3f464a4af71b1ca4227e7fa3f772deb84fc572f8b6f23661f733ffd079734"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:14:53.007188Z","signature_b64":"70BHoEoAzu66lrnWipVzdIV3UWsDt0JMqMI49LlgNqhOix3DRYhcPE548pyejnN6I4LQGzudodKY6KQ9nZIAAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"45ec7957b17863131b4a0791685eb0077daec241541b2427f16c9269b93b6a59","last_reissued_at":"2026-07-05T11:14:53.006676Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:14:53.006676Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Large Language Models to Diffusion Finetuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Edoardo Cetin, Tianyu Zhao, Yujin Tang","submitted_at":"2025-01-27T04:59:29Z","abstract_excerpt":"We propose a new finetuning method to provide pre-trained large language models (LMs) the ability to scale test-time compute through the diffusion framework. By increasing the number of diffusion steps, we show our finetuned models achieve monotonically increasing accuracy, directly translating to improved performance across downstream tasks. Furthermore, our finetuned models can expertly answer questions on specific topics by integrating powerful guidance techniques, and autonomously determine the compute required for a given problem by leveraging adaptive ODE solvers. Our method is universal"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.15781","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/2501.15781/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":"2501.15781","created_at":"2026-07-05T11:14:53.006741+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.15781v2","created_at":"2026-07-05T11:14:53.006741+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.15781","created_at":"2026-07-05T11:14:53.006741+00:00"},{"alias_kind":"pith_short_12","alias_value":"IXWHSV5RPBRR","created_at":"2026-07-05T11:14:53.006741+00:00"},{"alias_kind":"pith_short_16","alias_value":"IXWHSV5RPBRRGG2K","created_at":"2026-07-05T11:14:53.006741+00:00"},{"alias_kind":"pith_short_8","alias_value":"IXWHSV5R","created_at":"2026-07-05T11:14:53.006741+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/IXWHSV5RPBRRGG2KA6IWQXVQA5","json":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5.json","graph_json":"https://pith.science/api/pith-number/IXWHSV5RPBRRGG2KA6IWQXVQA5/graph.json","events_json":"https://pith.science/api/pith-number/IXWHSV5RPBRRGG2KA6IWQXVQA5/events.json","paper":"https://pith.science/paper/IXWHSV5R"},"agent_actions":{"view_html":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5","download_json":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5.json","view_paper":"https://pith.science/paper/IXWHSV5R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.15781&json=true","fetch_graph":"https://pith.science/api/pith-number/IXWHSV5RPBRRGG2KA6IWQXVQA5/graph.json","fetch_events":"https://pith.science/api/pith-number/IXWHSV5RPBRRGG2KA6IWQXVQA5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5/action/storage_attestation","attest_author":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5/action/author_attestation","sign_citation":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5/action/citation_signature","submit_replication":"https://pith.science/pith/IXWHSV5RPBRRGG2KA6IWQXVQA5/action/replication_record"}},"created_at":"2026-07-05T11:14:53.006741+00:00","updated_at":"2026-07-05T11:14:53.006741+00:00"}