{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VV5AR2DBFONM6ZHYSSJWN7UHAJ","short_pith_number":"pith:VV5AR2DB","schema_version":"1.0","canonical_sha256":"ad7a08e8612b9acf64f8949366fe87026d75125032a37498150f2173663afb60","source":{"kind":"arxiv","id":"2505.20297","version":1},"attestation_state":"computed","paper":{"title":"DiSA: Diffusion Step Annealing in Autoregressive Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Akshay Asthana, Jaskirat Singh, Liang Zheng, Ming Xu, Qinyu Zhao, Stephen Gould","submitted_at":"2025-05-26T17:59:57Z","abstract_excerpt":"An increasing number of autoregressive models, such as MAR, FlowAR, xAR, and Harmon adopt diffusion sampling to improve the quality of image generation. However, this strategy leads to low inference efficiency, because it usually takes 50 to 100 steps for diffusion to sample a token. This paper explores how to effectively address this issue. Our key motivation is that as more tokens are generated during the autoregressive process, subsequent tokens follow more constrained distributions and are easier to sample. To intuitively explain, if a model has generated part of a dog, the remaining token"},"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":"2505.20297","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-05-26T17:59:57Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"05feaf2f4b7433c977ae17a3b1e00030b96d4f8d08b9da66826240692122d769","abstract_canon_sha256":"153a473c7810a834f6103b53eedcf0720da376b4f086cf1e7bcfd42aebdbd9c6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:55.930954Z","signature_b64":"oBEzmXU1i43XwjxUCB4kjqFAXiAx39y4q/XHk37u3SDtoHiPUK1z0terkEbXxC3lyXbutOuRyVytMAF1o8DSAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ad7a08e8612b9acf64f8949366fe87026d75125032a37498150f2173663afb60","last_reissued_at":"2026-07-05T11:09:55.930409Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:55.930409Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DiSA: Diffusion Step Annealing in Autoregressive Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Akshay Asthana, Jaskirat Singh, Liang Zheng, Ming Xu, Qinyu Zhao, Stephen Gould","submitted_at":"2025-05-26T17:59:57Z","abstract_excerpt":"An increasing number of autoregressive models, such as MAR, FlowAR, xAR, and Harmon adopt diffusion sampling to improve the quality of image generation. However, this strategy leads to low inference efficiency, because it usually takes 50 to 100 steps for diffusion to sample a token. This paper explores how to effectively address this issue. Our key motivation is that as more tokens are generated during the autoregressive process, subsequent tokens follow more constrained distributions and are easier to sample. To intuitively explain, if a model has generated part of a dog, the remaining token"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.20297","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/2505.20297/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":"2505.20297","created_at":"2026-07-05T11:09:55.930470+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.20297v1","created_at":"2026-07-05T11:09:55.930470+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.20297","created_at":"2026-07-05T11:09:55.930470+00:00"},{"alias_kind":"pith_short_12","alias_value":"VV5AR2DBFONM","created_at":"2026-07-05T11:09:55.930470+00:00"},{"alias_kind":"pith_short_16","alias_value":"VV5AR2DBFONM6ZHY","created_at":"2026-07-05T11:09:55.930470+00:00"},{"alias_kind":"pith_short_8","alias_value":"VV5AR2DB","created_at":"2026-07-05T11:09:55.930470+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/VV5AR2DBFONM6ZHYSSJWN7UHAJ","json":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ.json","graph_json":"https://pith.science/api/pith-number/VV5AR2DBFONM6ZHYSSJWN7UHAJ/graph.json","events_json":"https://pith.science/api/pith-number/VV5AR2DBFONM6ZHYSSJWN7UHAJ/events.json","paper":"https://pith.science/paper/VV5AR2DB"},"agent_actions":{"view_html":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ","download_json":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ.json","view_paper":"https://pith.science/paper/VV5AR2DB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.20297&json=true","fetch_graph":"https://pith.science/api/pith-number/VV5AR2DBFONM6ZHYSSJWN7UHAJ/graph.json","fetch_events":"https://pith.science/api/pith-number/VV5AR2DBFONM6ZHYSSJWN7UHAJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ/action/storage_attestation","attest_author":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ/action/author_attestation","sign_citation":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ/action/citation_signature","submit_replication":"https://pith.science/pith/VV5AR2DBFONM6ZHYSSJWN7UHAJ/action/replication_record"}},"created_at":"2026-07-05T11:09:55.930470+00:00","updated_at":"2026-07-05T11:09:55.930470+00:00"}