{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CVJK6RMVS7LW6STFG436RVJE63","short_pith_number":"pith:CVJK6RMV","schema_version":"1.0","canonical_sha256":"1552af459597d76f4a653737e8d524f6feaf12b7828095f0d3d9eb1507b58ab2","source":{"kind":"arxiv","id":"2510.05356","version":2},"attestation_state":"computed","paper":{"title":"Mitigating Diffusion Model Hallucinations with Dynamic Guidance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Aggelina Chatziagapi, Alexandros Graikos, Dimitris Samaras, Grigorios G. Chrysos, Kostas Triaridis","submitted_at":"2025-10-06T20:31:13Z","abstract_excerpt":"Hallucinations in diffusion models are samples with structural inconsistencies that can emerge due to the excessive smoothing of the learned score function, which in turn leads to interpolations between modes of the data distribution. Since semantic interpolations are often desirable and contribute to sample diversity, we believe that a nuanced and targeted solution is required to address diffusion model hallucinations. In this work, we introduce Dynamic Guidance, which mitigates hallucinations by selectively sharpening the score function only along the pre-determined directions known to cause"},"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":"2510.05356","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-10-06T20:31:13Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"5601e71a1270cca8d7c303a4575cd746befdec73b7e3dfb286c7dae02c5b7b78","abstract_canon_sha256":"8cd9b9a48ce3ff611d5d21a567bbc089ed5227ee4f4e91e1929c14f464a5d666"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:04:39.461361Z","signature_b64":"KRPsvskvYe+XjH2BddviCyljwGVVYAg1VHBO3SkNfbcG0yzpStri7q6PbrRK8ZpSZ8Zvm62CfsQuBqt4dyohAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1552af459597d76f4a653737e8d524f6feaf12b7828095f0d3d9eb1507b58ab2","last_reissued_at":"2026-06-09T01:04:39.460872Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:04:39.460872Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mitigating Diffusion Model Hallucinations with Dynamic Guidance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Aggelina Chatziagapi, Alexandros Graikos, Dimitris Samaras, Grigorios G. Chrysos, Kostas Triaridis","submitted_at":"2025-10-06T20:31:13Z","abstract_excerpt":"Hallucinations in diffusion models are samples with structural inconsistencies that can emerge due to the excessive smoothing of the learned score function, which in turn leads to interpolations between modes of the data distribution. Since semantic interpolations are often desirable and contribute to sample diversity, we believe that a nuanced and targeted solution is required to address diffusion model hallucinations. In this work, we introduce Dynamic Guidance, which mitigates hallucinations by selectively sharpening the score function only along the pre-determined directions known to cause"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.05356","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/2510.05356/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":"2510.05356","created_at":"2026-06-09T01:04:39.460931+00:00"},{"alias_kind":"arxiv_version","alias_value":"2510.05356v2","created_at":"2026-06-09T01:04:39.460931+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.05356","created_at":"2026-06-09T01:04:39.460931+00:00"},{"alias_kind":"pith_short_12","alias_value":"CVJK6RMVS7LW","created_at":"2026-06-09T01:04:39.460931+00:00"},{"alias_kind":"pith_short_16","alias_value":"CVJK6RMVS7LW6STF","created_at":"2026-06-09T01:04:39.460931+00:00"},{"alias_kind":"pith_short_8","alias_value":"CVJK6RMV","created_at":"2026-06-09T01:04:39.460931+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2606.00377","citing_title":"Score-Control for Hallucination Reduction in Diffusion Models","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06831","citing_title":"Why DDIM Hallucinates More Than DDPM: A Theoretical Analysis of Reverse Dynamics","ref_index":36,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63","json":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63.json","graph_json":"https://pith.science/api/pith-number/CVJK6RMVS7LW6STFG436RVJE63/graph.json","events_json":"https://pith.science/api/pith-number/CVJK6RMVS7LW6STFG436RVJE63/events.json","paper":"https://pith.science/paper/CVJK6RMV"},"agent_actions":{"view_html":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63","download_json":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63.json","view_paper":"https://pith.science/paper/CVJK6RMV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2510.05356&json=true","fetch_graph":"https://pith.science/api/pith-number/CVJK6RMVS7LW6STFG436RVJE63/graph.json","fetch_events":"https://pith.science/api/pith-number/CVJK6RMVS7LW6STFG436RVJE63/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63/action/storage_attestation","attest_author":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63/action/author_attestation","sign_citation":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63/action/citation_signature","submit_replication":"https://pith.science/pith/CVJK6RMVS7LW6STFG436RVJE63/action/replication_record"}},"created_at":"2026-06-09T01:04:39.460931+00:00","updated_at":"2026-06-09T01:04:39.460931+00:00"}