{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VIEDBVOJVHXYZUSUKPEPYIPW32","short_pith_number":"pith:VIEDBVOJ","schema_version":"1.0","canonical_sha256":"aa0830d5c9a9ef8cd25453c8fc21f6de8a67e4d50879a83ffad638d01e9bb001","source":{"kind":"arxiv","id":"2504.10540","version":1},"attestation_state":"computed","paper":{"title":"AB-Cache: Training-Free Acceleration of Diffusion Models via Adams-Bashforth Cached Feature Reuse","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"stat.ML","authors_text":"Chengwei Zhang, Feng Zhao, Guojiang Shao, Jie Huang, Shengze Xu, Wenyi Zhang, Xiaodong Cun, Zhen Zou, Zichao Yu","submitted_at":"2025-04-13T08:29:58Z","abstract_excerpt":"Diffusion models have demonstrated remarkable success in generative tasks, yet their iterative denoising process results in slow inference, limiting their practicality. While existing acceleration methods exploit the well-known U-shaped similarity pattern between adjacent steps through caching mechanisms, they lack theoretical foundation and rely on simplistic computation reuse, often leading to performance degradation. In this work, we provide a theoretical understanding by analyzing the denoising process through the second-order Adams-Bashforth method, revealing a linear relationship between"},"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":"2504.10540","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-04-13T08:29:58Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"9721c645bac68bb0aa7bd738518368915839ae8e27627f41986ce85b467a3e1e","abstract_canon_sha256":"1cfa7be23f7f55fda5b5b0f1e0c1c6850b0a73b5103d8f12514fb736ae268822"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:49:19.229220Z","signature_b64":"O+e1bE5V5fmL8rRfUVjeJmWVuLyXCtvohdrtthS4/ERBA7j/QEuyLV546dcF1qJpZPV2SRxEBmMgzWC7Ae1wBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aa0830d5c9a9ef8cd25453c8fc21f6de8a67e4d50879a83ffad638d01e9bb001","last_reissued_at":"2026-07-05T10:49:19.228723Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:49:19.228723Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AB-Cache: Training-Free Acceleration of Diffusion Models via Adams-Bashforth Cached Feature Reuse","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"stat.ML","authors_text":"Chengwei Zhang, Feng Zhao, Guojiang Shao, Jie Huang, Shengze Xu, Wenyi Zhang, Xiaodong Cun, Zhen Zou, Zichao Yu","submitted_at":"2025-04-13T08:29:58Z","abstract_excerpt":"Diffusion models have demonstrated remarkable success in generative tasks, yet their iterative denoising process results in slow inference, limiting their practicality. While existing acceleration methods exploit the well-known U-shaped similarity pattern between adjacent steps through caching mechanisms, they lack theoretical foundation and rely on simplistic computation reuse, often leading to performance degradation. In this work, we provide a theoretical understanding by analyzing the denoising process through the second-order Adams-Bashforth method, revealing a linear relationship between"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.10540","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/2504.10540/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":"2504.10540","created_at":"2026-07-05T10:49:19.228788+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.10540v1","created_at":"2026-07-05T10:49:19.228788+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.10540","created_at":"2026-07-05T10:49:19.228788+00:00"},{"alias_kind":"pith_short_12","alias_value":"VIEDBVOJVHXY","created_at":"2026-07-05T10:49:19.228788+00:00"},{"alias_kind":"pith_short_16","alias_value":"VIEDBVOJVHXYZUSU","created_at":"2026-07-05T10:49:19.228788+00:00"},{"alias_kind":"pith_short_8","alias_value":"VIEDBVOJ","created_at":"2026-07-05T10:49:19.228788+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/VIEDBVOJVHXYZUSUKPEPYIPW32","json":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32.json","graph_json":"https://pith.science/api/pith-number/VIEDBVOJVHXYZUSUKPEPYIPW32/graph.json","events_json":"https://pith.science/api/pith-number/VIEDBVOJVHXYZUSUKPEPYIPW32/events.json","paper":"https://pith.science/paper/VIEDBVOJ"},"agent_actions":{"view_html":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32","download_json":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32.json","view_paper":"https://pith.science/paper/VIEDBVOJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.10540&json=true","fetch_graph":"https://pith.science/api/pith-number/VIEDBVOJVHXYZUSUKPEPYIPW32/graph.json","fetch_events":"https://pith.science/api/pith-number/VIEDBVOJVHXYZUSUKPEPYIPW32/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32/action/storage_attestation","attest_author":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32/action/author_attestation","sign_citation":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32/action/citation_signature","submit_replication":"https://pith.science/pith/VIEDBVOJVHXYZUSUKPEPYIPW32/action/replication_record"}},"created_at":"2026-07-05T10:49:19.228788+00:00","updated_at":"2026-07-05T10:49:19.228788+00:00"}