{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:QUJKBA3QPBQEFR5P3MJ52GXVDX","short_pith_number":"pith:QUJKBA3Q","schema_version":"1.0","canonical_sha256":"8512a08370786042c7afdb13dd1af51dd712e99aa09951eb41015755a72da192","source":{"kind":"arxiv","id":"2406.09408","version":3},"attestation_state":"computed","paper":{"title":"Data Attribution for Text-to-Image Models by Unlearning Synthesized Images","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Aaron Hertzmann, Alexei A. Efros, Jun-Yan Zhu, Richard Zhang, Sheng-Yu Wang","submitted_at":"2024-06-13T17:59:44Z","abstract_excerpt":"The goal of data attribution for text-to-image models is to identify the training images that most influence the generation of a new image. Influence is defined such that, for a given output, if a model is retrained from scratch without the most influential images, the model would fail to reproduce the same output. Unfortunately, directly searching for these influential images is computationally infeasible, since it would require repeatedly retraining models from scratch. In our work, we propose an efficient data attribution method by simulating unlearning the synthesized image. We achieve thi"},"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":"2406.09408","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-06-13T17:59:44Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"e9c9c6a1fe2431798f8e1d71e707185e79d52be1cb505b7babf0395b488eceb9","abstract_canon_sha256":"32b7adf98c1d5a469c80e28dc80dc63adc2d3e9913be3a236aaaaddfb383468c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:17:08.895357Z","signature_b64":"o3mhZuTVLzk/Nxw4VmunAVz1Ye6eplX6v2bT6mMpyaCmS3q05vuiMrSsa8RWUiq7XOpcdUGbo19uFtYNuYi+Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8512a08370786042c7afdb13dd1af51dd712e99aa09951eb41015755a72da192","last_reissued_at":"2026-07-05T10:17:08.894814Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:17:08.894814Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Data Attribution for Text-to-Image Models by Unlearning Synthesized Images","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Aaron Hertzmann, Alexei A. Efros, Jun-Yan Zhu, Richard Zhang, Sheng-Yu Wang","submitted_at":"2024-06-13T17:59:44Z","abstract_excerpt":"The goal of data attribution for text-to-image models is to identify the training images that most influence the generation of a new image. Influence is defined such that, for a given output, if a model is retrained from scratch without the most influential images, the model would fail to reproduce the same output. Unfortunately, directly searching for these influential images is computationally infeasible, since it would require repeatedly retraining models from scratch. In our work, we propose an efficient data attribution method by simulating unlearning the synthesized image. We achieve thi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.09408","kind":"arxiv","version":3},"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/2406.09408/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":"2406.09408","created_at":"2026-07-05T10:17:08.894869+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.09408v3","created_at":"2026-07-05T10:17:08.894869+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.09408","created_at":"2026-07-05T10:17:08.894869+00:00"},{"alias_kind":"pith_short_12","alias_value":"QUJKBA3QPBQE","created_at":"2026-07-05T10:17:08.894869+00:00"},{"alias_kind":"pith_short_16","alias_value":"QUJKBA3QPBQEFR5P","created_at":"2026-07-05T10:17:08.894869+00:00"},{"alias_kind":"pith_short_8","alias_value":"QUJKBA3Q","created_at":"2026-07-05T10:17:08.894869+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.14432","citing_title":"IntroStyle: Training-Free Introspective Style Attribution using Diffusion Features","ref_index":44,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX","json":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX.json","graph_json":"https://pith.science/api/pith-number/QUJKBA3QPBQEFR5P3MJ52GXVDX/graph.json","events_json":"https://pith.science/api/pith-number/QUJKBA3QPBQEFR5P3MJ52GXVDX/events.json","paper":"https://pith.science/paper/QUJKBA3Q"},"agent_actions":{"view_html":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX","download_json":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX.json","view_paper":"https://pith.science/paper/QUJKBA3Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.09408&json=true","fetch_graph":"https://pith.science/api/pith-number/QUJKBA3QPBQEFR5P3MJ52GXVDX/graph.json","fetch_events":"https://pith.science/api/pith-number/QUJKBA3QPBQEFR5P3MJ52GXVDX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX/action/storage_attestation","attest_author":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX/action/author_attestation","sign_citation":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX/action/citation_signature","submit_replication":"https://pith.science/pith/QUJKBA3QPBQEFR5P3MJ52GXVDX/action/replication_record"}},"created_at":"2026-07-05T10:17:08.894869+00:00","updated_at":"2026-07-05T10:17:08.894869+00:00"}