{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:JYIC3N5FB5CLEKHBPTYT7OTKMZ","short_pith_number":"pith:JYIC3N5F","schema_version":"1.0","canonical_sha256":"4e102db7a50f44b228e17cf13fba6a667c39d8f7730bfefd9f4e82c24d420cdd","source":{"kind":"arxiv","id":"2606.28401","version":1},"attestation_state":"computed","paper":{"title":"Vision-driven Preference Synthesis for Mitigating Hallucinations in VLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Jongheon Jeong, Yunhun Nam","submitted_at":"2026-06-24T11:06:22Z","abstract_excerpt":"Vision-Language Models (VLMs) have shown strong performance in visual understanding, yet they still suffer from hallucinations, generating content that is not grounded in the image. Preference alignment is a promising approach to improve visual faithfulness, but its success depends heavily on how preference pairs are constructed. Existing methods exhibit two key limitations; (a) intervention-based methods often introduce significant deviation from the policy distribution, and (b) sampling-based methods often underuse visual information during the construction. In this paper, we propose ViPSy ("},"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":"2606.28401","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-24T11:06:22Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"c036f2c6c904d7953be1fadbcd3acd133d2446d429550de13080427441964f8f","abstract_canon_sha256":"6762f3597784244acc18aabb16d273a16857f338a423ea96846d905a7bbacc1c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T00:15:13.071127Z","signature_b64":"maBEe7sXGRO4Wx22YRqJUjexzivMupGp1bsHmFXrKD0Wi6T9qsKu8MrM0YB+OFqp9Bz3z65z03BH7ZyI1V4VBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4e102db7a50f44b228e17cf13fba6a667c39d8f7730bfefd9f4e82c24d420cdd","last_reissued_at":"2026-06-30T00:15:13.070732Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T00:15:13.070732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Vision-driven Preference Synthesis for Mitigating Hallucinations in VLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Jongheon Jeong, Yunhun Nam","submitted_at":"2026-06-24T11:06:22Z","abstract_excerpt":"Vision-Language Models (VLMs) have shown strong performance in visual understanding, yet they still suffer from hallucinations, generating content that is not grounded in the image. Preference alignment is a promising approach to improve visual faithfulness, but its success depends heavily on how preference pairs are constructed. Existing methods exhibit two key limitations; (a) intervention-based methods often introduce significant deviation from the policy distribution, and (b) sampling-based methods often underuse visual information during the construction. In this paper, we propose ViPSy ("},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.28401","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/2606.28401/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":"2606.28401","created_at":"2026-06-30T00:15:13.070791+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.28401v1","created_at":"2026-06-30T00:15:13.070791+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.28401","created_at":"2026-06-30T00:15:13.070791+00:00"},{"alias_kind":"pith_short_12","alias_value":"JYIC3N5FB5CL","created_at":"2026-06-30T00:15:13.070791+00:00"},{"alias_kind":"pith_short_16","alias_value":"JYIC3N5FB5CLEKHB","created_at":"2026-06-30T00:15:13.070791+00:00"},{"alias_kind":"pith_short_8","alias_value":"JYIC3N5F","created_at":"2026-06-30T00:15:13.070791+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/JYIC3N5FB5CLEKHBPTYT7OTKMZ","json":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ.json","graph_json":"https://pith.science/api/pith-number/JYIC3N5FB5CLEKHBPTYT7OTKMZ/graph.json","events_json":"https://pith.science/api/pith-number/JYIC3N5FB5CLEKHBPTYT7OTKMZ/events.json","paper":"https://pith.science/paper/JYIC3N5F"},"agent_actions":{"view_html":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ","download_json":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ.json","view_paper":"https://pith.science/paper/JYIC3N5F","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.28401&json=true","fetch_graph":"https://pith.science/api/pith-number/JYIC3N5FB5CLEKHBPTYT7OTKMZ/graph.json","fetch_events":"https://pith.science/api/pith-number/JYIC3N5FB5CLEKHBPTYT7OTKMZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ/action/storage_attestation","attest_author":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ/action/author_attestation","sign_citation":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ/action/citation_signature","submit_replication":"https://pith.science/pith/JYIC3N5FB5CLEKHBPTYT7OTKMZ/action/replication_record"}},"created_at":"2026-06-30T00:15:13.070791+00:00","updated_at":"2026-06-30T00:15:13.070791+00:00"}