{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DOOWQ4BPSQTPX2HQH36B5NR2N3","short_pith_number":"pith:DOOWQ4BP","schema_version":"1.0","canonical_sha256":"1b9d68702f9426fbe8f03efc1eb63a6ed2505055652cea467c9f68068b57260f","source":{"kind":"arxiv","id":"2505.17352","version":2},"attestation_state":"computed","paper":{"title":"Alignment and Safety of Diffusion Models via Reinforcement Learning and Reward Modeling: A Survey","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ankita Kushwaha, Kiran Ravish, Pawan Kumar, Preeti Lamba","submitted_at":"2025-05-23T00:08:49Z","abstract_excerpt":"Diffusion models have become a central paradigm for image and multimodal generation, yet their deployment raises persistent questions about alignment, safety, preference satisfaction, and robustness to misuse. This survey reviews recent progress on aligning text-to-image diffusion models through reinforcement learning, reward modeling, preference optimization, and safety-specific fine-tuning. We organize the literature along five axes: the source of feedback, the form of the reward or preference signal, the optimization mechanism, the treatment of distribution shift and reward overoptimization"},"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":"2505.17352","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-05-23T00:08:49Z","cross_cats_sorted":[],"title_canon_sha256":"5b5f8311475828b31c6cd9880bfe01d5f5011f460938092523be5a25f369e79e","abstract_canon_sha256":"822cdc021426df8a23d2928052bab6f7262bb89c01130266cff405920ce49922"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:05.165715Z","signature_b64":"X79/oDXPzmG4aidx6Iy+U1ZwgwEiFSI5xq/+Q48bt7gF4emRjAvHuYmtuhX4jbHx0/MD4wWD/Au1PRXkVdx+Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1b9d68702f9426fbe8f03efc1eb63a6ed2505055652cea467c9f68068b57260f","last_reissued_at":"2026-05-20T00:02:05.164986Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:05.164986Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Alignment and Safety of Diffusion Models via Reinforcement Learning and Reward Modeling: A Survey","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ankita Kushwaha, Kiran Ravish, Pawan Kumar, Preeti Lamba","submitted_at":"2025-05-23T00:08:49Z","abstract_excerpt":"Diffusion models have become a central paradigm for image and multimodal generation, yet their deployment raises persistent questions about alignment, safety, preference satisfaction, and robustness to misuse. This survey reviews recent progress on aligning text-to-image diffusion models through reinforcement learning, reward modeling, preference optimization, and safety-specific fine-tuning. We organize the literature along five axes: the source of feedback, the form of the reward or preference signal, the optimization mechanism, the treatment of distribution shift and reward overoptimization"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.17352","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/2505.17352/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":"2505.17352","created_at":"2026-05-20T00:02:05.165099+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.17352v2","created_at":"2026-05-20T00:02:05.165099+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.17352","created_at":"2026-05-20T00:02:05.165099+00:00"},{"alias_kind":"pith_short_12","alias_value":"DOOWQ4BPSQTP","created_at":"2026-05-20T00:02:05.165099+00:00"},{"alias_kind":"pith_short_16","alias_value":"DOOWQ4BPSQTPX2HQ","created_at":"2026-05-20T00:02:05.165099+00:00"},{"alias_kind":"pith_short_8","alias_value":"DOOWQ4BP","created_at":"2026-05-20T00:02:05.165099+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.18719","citing_title":"SafeDiffusion-R1: Online Reward Steering for Safe Diffusion Post-Training","ref_index":86,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3","json":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3.json","graph_json":"https://pith.science/api/pith-number/DOOWQ4BPSQTPX2HQH36B5NR2N3/graph.json","events_json":"https://pith.science/api/pith-number/DOOWQ4BPSQTPX2HQH36B5NR2N3/events.json","paper":"https://pith.science/paper/DOOWQ4BP"},"agent_actions":{"view_html":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3","download_json":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3.json","view_paper":"https://pith.science/paper/DOOWQ4BP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.17352&json=true","fetch_graph":"https://pith.science/api/pith-number/DOOWQ4BPSQTPX2HQH36B5NR2N3/graph.json","fetch_events":"https://pith.science/api/pith-number/DOOWQ4BPSQTPX2HQH36B5NR2N3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3/action/storage_attestation","attest_author":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3/action/author_attestation","sign_citation":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3/action/citation_signature","submit_replication":"https://pith.science/pith/DOOWQ4BPSQTPX2HQH36B5NR2N3/action/replication_record"}},"created_at":"2026-05-20T00:02:05.165099+00:00","updated_at":"2026-05-20T00:02:05.165099+00:00"}