{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NXPSBOJQDTXB2YMHFZOLBFUWCM","short_pith_number":"pith:NXPSBOJQ","schema_version":"1.0","canonical_sha256":"6ddf20b9301cee1d61872e5cb09696133c7452e43c4cfa1443615e35b70a4eb5","source":{"kind":"arxiv","id":"2602.19149","version":3},"attestation_state":"computed","paper":{"title":"ReVision : A Post-Hoc, Vision-Based Technique for Replacing Unacceptable Concepts in Image Generation Pipeline","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Aashima Sharma, Gurjot Singh, Maninder Singh, Prabhjot Singh, Ryan Ko","submitted_at":"2026-02-22T12:30:01Z","abstract_excerpt":"Image-generative models are widely deployed across industries. Recent studies show that they can be exploited to produce unacceptable content. Existing mitigation strategies rely on prompt filtering and safety-aware training, both of which can be bypassed and often degrade generative quality. In this work, we propose ReVision, a training-free, prompt-based, post-hoc safety framework for image-generation pipeline. ReVision acts as a post-generation safeguard by analyzing generated images and selectively editing unsafe concepts without altering the underlying generator. Prior post-hoc editing me"},"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":"2602.19149","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-02-22T12:30:01Z","cross_cats_sorted":[],"title_canon_sha256":"fd958c2b048d7bdf48549d35605efb6ea4d4725b61fed2b75fe01e89757d96e2","abstract_canon_sha256":"7ca92fb9a6f6f28592859cf95fcd1b9c323ba4c962e0d78f8f4745a6520c3e13"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T00:21:34.402556Z","signature_b64":"pwLNeD63zEbKavpgzUO8nKySw0yS77eo+9mMosYIJHt1eVZ9OFgpCGeVM9o0rs6LzsxxFDxGalsGr24ZgZimCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6ddf20b9301cee1d61872e5cb09696133c7452e43c4cfa1443615e35b70a4eb5","last_reissued_at":"2026-07-28T00:21:34.401627Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T00:21:34.401627Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReVision : A Post-Hoc, Vision-Based Technique for Replacing Unacceptable Concepts in Image Generation Pipeline","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Aashima Sharma, Gurjot Singh, Maninder Singh, Prabhjot Singh, Ryan Ko","submitted_at":"2026-02-22T12:30:01Z","abstract_excerpt":"Image-generative models are widely deployed across industries. Recent studies show that they can be exploited to produce unacceptable content. Existing mitigation strategies rely on prompt filtering and safety-aware training, both of which can be bypassed and often degrade generative quality. In this work, we propose ReVision, a training-free, prompt-based, post-hoc safety framework for image-generation pipeline. ReVision acts as a post-generation safeguard by analyzing generated images and selectively editing unsafe concepts without altering the underlying generator. Prior post-hoc editing me"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.19149","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/2602.19149/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":"2602.19149","created_at":"2026-07-28T00:21:34.402056+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.19149v3","created_at":"2026-07-28T00:21:34.402056+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.19149","created_at":"2026-07-28T00:21:34.402056+00:00"},{"alias_kind":"pith_short_12","alias_value":"NXPSBOJQDTXB","created_at":"2026-07-28T00:21:34.402056+00:00"},{"alias_kind":"pith_short_16","alias_value":"NXPSBOJQDTXB2YMH","created_at":"2026-07-28T00:21:34.402056+00:00"},{"alias_kind":"pith_short_8","alias_value":"NXPSBOJQ","created_at":"2026-07-28T00:21:34.402056+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.30049","citing_title":"Robust and Generalizable Safety Steering for Text-to-Image Diffusion Transformers","ref_index":41,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM","json":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM.json","graph_json":"https://pith.science/api/pith-number/NXPSBOJQDTXB2YMHFZOLBFUWCM/graph.json","events_json":"https://pith.science/api/pith-number/NXPSBOJQDTXB2YMHFZOLBFUWCM/events.json","paper":"https://pith.science/paper/NXPSBOJQ"},"agent_actions":{"view_html":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM","download_json":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM.json","view_paper":"https://pith.science/paper/NXPSBOJQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.19149&json=true","fetch_graph":"https://pith.science/api/pith-number/NXPSBOJQDTXB2YMHFZOLBFUWCM/graph.json","fetch_events":"https://pith.science/api/pith-number/NXPSBOJQDTXB2YMHFZOLBFUWCM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM/action/storage_attestation","attest_author":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM/action/author_attestation","sign_citation":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM/action/citation_signature","submit_replication":"https://pith.science/pith/NXPSBOJQDTXB2YMHFZOLBFUWCM/action/replication_record"}},"created_at":"2026-07-28T00:21:34.402056+00:00","updated_at":"2026-07-28T00:21:34.402056+00:00"}