{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:SL6TFGIBMA7ZAK3JKOYH4V4N7C","short_pith_number":"pith:SL6TFGIB","schema_version":"1.0","canonical_sha256":"92fd329901603f902b6953b07e578df894b31699bbbf61de22fadb51baa0242d","source":{"kind":"arxiv","id":"2412.10891","version":2},"attestation_state":"computed","paper":{"title":"Zigzag Diffusion Sampling: Diffusion Models Can Self-Improve via Self-Reflection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Haoyi Xiong, Lichen Bai, Shitong Shao, Zeke Xie, Zhiqiang Xu, Zikai Zhou, Zipeng Qi","submitted_at":"2024-12-14T16:42:41Z","abstract_excerpt":"Diffusion models, the most popular generative paradigm so far, can inject conditional information into the generation path to guide the latent towards desired directions. However, existing text-to-image diffusion models often fail to maintain high image quality and high prompt-image alignment for those challenging prompts. To mitigate this issue and enhance existing pretrained diffusion models, we mainly made three contributions in this paper. First, we propose diffusion self-reflection that alternately performs denoising and inversion and demonstrate that such diffusion self-reflection can le"},"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":"2412.10891","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-14T16:42:41Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4913d2e15192d4bfac023084f546b499a6048309fa949cf40626c5407889ea9f","abstract_canon_sha256":"3f712fc470e2a5714fcdd87a5ab8c9c7df3135b52df65cb3a4c91446fb943d23"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:50:08.951981Z","signature_b64":"PNFc1MIWZH4+7IUJDnDBbEW34CBvi+4AFZFemP05UunvFalW0Vz+NJuR3sNCErpUvZMiOweQzPlajv2E/NMiDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"92fd329901603f902b6953b07e578df894b31699bbbf61de22fadb51baa0242d","last_reissued_at":"2026-07-05T09:50:08.951493Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:50:08.951493Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Zigzag Diffusion Sampling: Diffusion Models Can Self-Improve via Self-Reflection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Haoyi Xiong, Lichen Bai, Shitong Shao, Zeke Xie, Zhiqiang Xu, Zikai Zhou, Zipeng Qi","submitted_at":"2024-12-14T16:42:41Z","abstract_excerpt":"Diffusion models, the most popular generative paradigm so far, can inject conditional information into the generation path to guide the latent towards desired directions. However, existing text-to-image diffusion models often fail to maintain high image quality and high prompt-image alignment for those challenging prompts. To mitigate this issue and enhance existing pretrained diffusion models, we mainly made three contributions in this paper. First, we propose diffusion self-reflection that alternately performs denoising and inversion and demonstrate that such diffusion self-reflection can le"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.10891","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/2412.10891/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":"2412.10891","created_at":"2026-07-05T09:50:08.951561+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.10891v2","created_at":"2026-07-05T09:50:08.951561+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.10891","created_at":"2026-07-05T09:50:08.951561+00:00"},{"alias_kind":"pith_short_12","alias_value":"SL6TFGIBMA7Z","created_at":"2026-07-05T09:50:08.951561+00:00"},{"alias_kind":"pith_short_16","alias_value":"SL6TFGIBMA7ZAK3J","created_at":"2026-07-05T09:50:08.951561+00:00"},{"alias_kind":"pith_short_8","alias_value":"SL6TFGIB","created_at":"2026-07-05T09:50:08.951561+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.16054","citing_title":"Ada-Diffuser: Latent-Aware Adaptive Diffusion for Decision-Making","ref_index":267,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23536","citing_title":"$Z^2$-Sampling: Zero-Cost Zigzag Trajectories for Semantic Alignment in Diffusion Models","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04646","citing_title":"Training-Free Refinement of Flow Matching with Divergence-based Sampling","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C","json":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C.json","graph_json":"https://pith.science/api/pith-number/SL6TFGIBMA7ZAK3JKOYH4V4N7C/graph.json","events_json":"https://pith.science/api/pith-number/SL6TFGIBMA7ZAK3JKOYH4V4N7C/events.json","paper":"https://pith.science/paper/SL6TFGIB"},"agent_actions":{"view_html":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C","download_json":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C.json","view_paper":"https://pith.science/paper/SL6TFGIB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.10891&json=true","fetch_graph":"https://pith.science/api/pith-number/SL6TFGIBMA7ZAK3JKOYH4V4N7C/graph.json","fetch_events":"https://pith.science/api/pith-number/SL6TFGIBMA7ZAK3JKOYH4V4N7C/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C/action/storage_attestation","attest_author":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C/action/author_attestation","sign_citation":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C/action/citation_signature","submit_replication":"https://pith.science/pith/SL6TFGIBMA7ZAK3JKOYH4V4N7C/action/replication_record"}},"created_at":"2026-07-05T09:50:08.951561+00:00","updated_at":"2026-07-05T09:50:08.951561+00:00"}