{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BJXBGJKXAJKEUNOAHWXB3W6BOQ","short_pith_number":"pith:BJXBGJKX","schema_version":"1.0","canonical_sha256":"0a6e13255702544a35c03dae1ddbc17425617f47443b48fc8997bbc8c040ec71","source":{"kind":"arxiv","id":"2607.00817","version":1},"attestation_state":"computed","paper":{"title":"Training-Free Debiasing of Diffusion Models via CLIP-Guided Denoising Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dain Kim, Jinseo Kim, Sungyong Baik","submitted_at":"2026-07-01T11:41:40Z","abstract_excerpt":"Text-to-image diffusion models achieve impressive visual quality, yet demographic bias remains a challenge, as neutral prompts consistently produce stereotypical representations across gender and race. Existing approaches remain limited by costly retraining or by inference-time interventions that often degrade image quality and semantic alignment. We propose Text Embedding Steering (TES), a training-free framework that mitigates demographic bias by directly optimizing conditional text embeddings during the diffusion process. We show that a two-stage strategy - early-stage global alignment foll"},"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":"2607.00817","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-07-01T11:41:40Z","cross_cats_sorted":[],"title_canon_sha256":"65b66c3d97e8c73913a8d91ebc2e39e9cca4e4515c15c8ebfd6da6d342f7ff58","abstract_canon_sha256":"463a93481e4972b8b76994fb83182c9c6a7f37bb52e9f18db86757f81d43733c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-02T01:17:56.192594Z","signature_b64":"Ixj1nHVsoqxX9jflBZhNse+x4ckF+biCj8VSdwd1v7DVeQ6ecpWNLpoRaxy9Fea/uy2eInRZfX1zYQumir4mBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0a6e13255702544a35c03dae1ddbc17425617f47443b48fc8997bbc8c040ec71","last_reissued_at":"2026-07-02T01:17:56.192183Z","signature_status":"signed_v1","first_computed_at":"2026-07-02T01:17:56.192183Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training-Free Debiasing of Diffusion Models via CLIP-Guided Denoising Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dain Kim, Jinseo Kim, Sungyong Baik","submitted_at":"2026-07-01T11:41:40Z","abstract_excerpt":"Text-to-image diffusion models achieve impressive visual quality, yet demographic bias remains a challenge, as neutral prompts consistently produce stereotypical representations across gender and race. Existing approaches remain limited by costly retraining or by inference-time interventions that often degrade image quality and semantic alignment. We propose Text Embedding Steering (TES), a training-free framework that mitigates demographic bias by directly optimizing conditional text embeddings during the diffusion process. We show that a two-stage strategy - early-stage global alignment foll"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.00817","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/2607.00817/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":"2607.00817","created_at":"2026-07-02T01:17:56.192244+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.00817v1","created_at":"2026-07-02T01:17:56.192244+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.00817","created_at":"2026-07-02T01:17:56.192244+00:00"},{"alias_kind":"pith_short_12","alias_value":"BJXBGJKXAJKE","created_at":"2026-07-02T01:17:56.192244+00:00"},{"alias_kind":"pith_short_16","alias_value":"BJXBGJKXAJKEUNOA","created_at":"2026-07-02T01:17:56.192244+00:00"},{"alias_kind":"pith_short_8","alias_value":"BJXBGJKX","created_at":"2026-07-02T01:17:56.192244+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/BJXBGJKXAJKEUNOAHWXB3W6BOQ","json":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ.json","graph_json":"https://pith.science/api/pith-number/BJXBGJKXAJKEUNOAHWXB3W6BOQ/graph.json","events_json":"https://pith.science/api/pith-number/BJXBGJKXAJKEUNOAHWXB3W6BOQ/events.json","paper":"https://pith.science/paper/BJXBGJKX"},"agent_actions":{"view_html":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ","download_json":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ.json","view_paper":"https://pith.science/paper/BJXBGJKX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.00817&json=true","fetch_graph":"https://pith.science/api/pith-number/BJXBGJKXAJKEUNOAHWXB3W6BOQ/graph.json","fetch_events":"https://pith.science/api/pith-number/BJXBGJKXAJKEUNOAHWXB3W6BOQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ/action/storage_attestation","attest_author":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ/action/author_attestation","sign_citation":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ/action/citation_signature","submit_replication":"https://pith.science/pith/BJXBGJKXAJKEUNOAHWXB3W6BOQ/action/replication_record"}},"created_at":"2026-07-02T01:17:56.192244+00:00","updated_at":"2026-07-02T01:17:56.192244+00:00"}