{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:S6T76EOVJAJN72R4F7ZRM6AV6V","short_pith_number":"pith:S6T76EOV","schema_version":"1.0","canonical_sha256":"97a7ff11d54812dfea3c2ff3167815f546a6539c1ce9d25cbb91ce311b416534","source":{"kind":"arxiv","id":"2410.18013","version":2},"attestation_state":"computed","paper":{"title":"Scalable Ranked Preference Optimization for Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Anil Kag, Huseyin Coskun, Jian Ren, Sergey Tulyakov, Shyamgopal Karthik, Zeynep Akata","submitted_at":"2024-10-23T16:42:56Z","abstract_excerpt":"Direct Preference Optimization (DPO) has emerged as a powerful approach to align text-to-image (T2I) models with human feedback. Unfortunately, successful application of DPO to T2I models requires a huge amount of resources to collect and label large-scale datasets, e.g., millions of generated paired images annotated with human preferences. In addition, these human preference datasets can get outdated quickly as the rapid improvements of T2I models lead to higher quality images. In this work, we investigate a scalable approach for collecting large-scale and fully synthetic datasets for DPO tra"},"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":"2410.18013","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-10-23T16:42:56Z","cross_cats_sorted":[],"title_canon_sha256":"67c1459df3d3953fbd6ef6930dcf071b3e08bd17f8902352a2f629e00ecca9f4","abstract_canon_sha256":"bb3fb8747d1632a7a1d78d88458160f6953b16dc78d119ff5a78fdabc9337131"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:28:35.190892Z","signature_b64":"mW9D8f9fcHfibuV+fg1mHc/KSNfGXZVm9MziMltLSZRMe2gRUY0QxM3i/bYKr/goLPxgJyKv0QohRKTdlRJHDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"97a7ff11d54812dfea3c2ff3167815f546a6539c1ce9d25cbb91ce311b416534","last_reissued_at":"2026-07-05T09:28:35.190480Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:28:35.190480Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scalable Ranked Preference Optimization for Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Anil Kag, Huseyin Coskun, Jian Ren, Sergey Tulyakov, Shyamgopal Karthik, Zeynep Akata","submitted_at":"2024-10-23T16:42:56Z","abstract_excerpt":"Direct Preference Optimization (DPO) has emerged as a powerful approach to align text-to-image (T2I) models with human feedback. Unfortunately, successful application of DPO to T2I models requires a huge amount of resources to collect and label large-scale datasets, e.g., millions of generated paired images annotated with human preferences. In addition, these human preference datasets can get outdated quickly as the rapid improvements of T2I models lead to higher quality images. In this work, we investigate a scalable approach for collecting large-scale and fully synthetic datasets for DPO tra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.18013","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/2410.18013/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":"2410.18013","created_at":"2026-07-05T09:28:35.190536+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.18013v2","created_at":"2026-07-05T09:28:35.190536+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.18013","created_at":"2026-07-05T09:28:35.190536+00:00"},{"alias_kind":"pith_short_12","alias_value":"S6T76EOVJAJN","created_at":"2026-07-05T09:28:35.190536+00:00"},{"alias_kind":"pith_short_16","alias_value":"S6T76EOVJAJN72R4","created_at":"2026-07-05T09:28:35.190536+00:00"},{"alias_kind":"pith_short_8","alias_value":"S6T76EOV","created_at":"2026-07-05T09:28:35.190536+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.28615","citing_title":"Compositional Text-to-Image Generation Via Region-aware Bimodal Direct Preference Optimization","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09433","citing_title":"Offline Preference Optimization for Rectified Flow with Noise-Tracked Pairs","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07427","citing_title":"Personalizing Text-to-Image Generation to Individual Taste","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V","json":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V.json","graph_json":"https://pith.science/api/pith-number/S6T76EOVJAJN72R4F7ZRM6AV6V/graph.json","events_json":"https://pith.science/api/pith-number/S6T76EOVJAJN72R4F7ZRM6AV6V/events.json","paper":"https://pith.science/paper/S6T76EOV"},"agent_actions":{"view_html":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V","download_json":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V.json","view_paper":"https://pith.science/paper/S6T76EOV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.18013&json=true","fetch_graph":"https://pith.science/api/pith-number/S6T76EOVJAJN72R4F7ZRM6AV6V/graph.json","fetch_events":"https://pith.science/api/pith-number/S6T76EOVJAJN72R4F7ZRM6AV6V/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V/action/storage_attestation","attest_author":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V/action/author_attestation","sign_citation":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V/action/citation_signature","submit_replication":"https://pith.science/pith/S6T76EOVJAJN72R4F7ZRM6AV6V/action/replication_record"}},"created_at":"2026-07-05T09:28:35.190536+00:00","updated_at":"2026-07-05T09:28:35.190536+00:00"}