{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:EJ7VEY7QPGKBP3ZQEH2PCA45ZQ","short_pith_number":"pith:EJ7VEY7Q","schema_version":"1.0","canonical_sha256":"227f5263f0799417ef3021f4f1039dcc37634c899dd9bdb3f85560af855e5af6","source":{"kind":"arxiv","id":"2506.09376","version":1},"attestation_state":"computed","paper":{"title":"Revisiting Diffusion Models: From Generative Pre-training to One-Step Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bowen Zheng, Tianming Yang","submitted_at":"2025-06-11T03:55:26Z","abstract_excerpt":"Diffusion distillation is a widely used technique to reduce the sampling cost of diffusion models, yet it often requires extensive training, and the student performance tends to be degraded. Recent studies show that incorporating a GAN objective may alleviate these issues, yet the underlying mechanism remains unclear. In this work, we first identify a key limitation of distillation: mismatched step sizes and parameter numbers between the teacher and the student model lead them to converge to different local minima, rendering direct imitation suboptimal. We further demonstrate that a standalone"},"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":"2506.09376","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-11T03:55:26Z","cross_cats_sorted":[],"title_canon_sha256":"571cb616ec129f92bea48fa2a218b932b51d43e210c867476508e2fe33fc4620","abstract_canon_sha256":"66b88518184dfdcb3162e0f393b0385e60783244c1c8f93daa0e3db6beb5b6be"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:36.019876Z","signature_b64":"wHvEfZAmwZ8xTswQmeAr/t9LxQWt7pNjZtqk4h5FM7HpMpBMKiloNjf4IKxI0aRcUSjU0i7s/PawZ/LflvYrDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"227f5263f0799417ef3021f4f1039dcc37634c899dd9bdb3f85560af855e5af6","last_reissued_at":"2026-07-05T11:19:36.019405Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:36.019405Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Revisiting Diffusion Models: From Generative Pre-training to One-Step Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bowen Zheng, Tianming Yang","submitted_at":"2025-06-11T03:55:26Z","abstract_excerpt":"Diffusion distillation is a widely used technique to reduce the sampling cost of diffusion models, yet it often requires extensive training, and the student performance tends to be degraded. Recent studies show that incorporating a GAN objective may alleviate these issues, yet the underlying mechanism remains unclear. In this work, we first identify a key limitation of distillation: mismatched step sizes and parameter numbers between the teacher and the student model lead them to converge to different local minima, rendering direct imitation suboptimal. We further demonstrate that a standalone"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.09376","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/2506.09376/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":"2506.09376","created_at":"2026-07-05T11:19:36.019465+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.09376v1","created_at":"2026-07-05T11:19:36.019465+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.09376","created_at":"2026-07-05T11:19:36.019465+00:00"},{"alias_kind":"pith_short_12","alias_value":"EJ7VEY7QPGKB","created_at":"2026-07-05T11:19:36.019465+00:00"},{"alias_kind":"pith_short_16","alias_value":"EJ7VEY7QPGKBP3ZQ","created_at":"2026-07-05T11:19:36.019465+00:00"},{"alias_kind":"pith_short_8","alias_value":"EJ7VEY7Q","created_at":"2026-07-05T11:19:36.019465+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.13713","citing_title":"Learning to Optimize Radiotherapy Plans via Fluence Maps Diffusion Model Generation and LSTM-based Optimization","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ","json":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ.json","graph_json":"https://pith.science/api/pith-number/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/graph.json","events_json":"https://pith.science/api/pith-number/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/events.json","paper":"https://pith.science/paper/EJ7VEY7Q"},"agent_actions":{"view_html":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ","download_json":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ.json","view_paper":"https://pith.science/paper/EJ7VEY7Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.09376&json=true","fetch_graph":"https://pith.science/api/pith-number/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/graph.json","fetch_events":"https://pith.science/api/pith-number/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/action/storage_attestation","attest_author":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/action/author_attestation","sign_citation":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/action/citation_signature","submit_replication":"https://pith.science/pith/EJ7VEY7QPGKBP3ZQEH2PCA45ZQ/action/replication_record"}},"created_at":"2026-07-05T11:19:36.019465+00:00","updated_at":"2026-07-05T11:19:36.019465+00:00"}