{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:IXEZ2DTHTKSS3U5IZRVCM5Q7KG","short_pith_number":"pith:IXEZ2DTH","schema_version":"1.0","canonical_sha256":"45c99d0e679aa52dd3a8cc6a26761f5197e3835541385a743db5f3cecd9dc46b","source":{"kind":"arxiv","id":"2605.16411","version":1},"attestation_state":"computed","paper":{"title":"Reducing Hallucination in Vision-Language Models via Stage-wise Preference Optimization under Distribution Shift","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.DB","cs.LG"],"primary_cat":"cs.CV","authors_text":"Qinwu Xu","submitted_at":"2026-05-13T15:37:51Z","abstract_excerpt":"Hallucination remains a fundamental challenge in vision-language models (VLMs), where autoregressive generation may produce linguistically plausible yet physically inconsistent or visually ungrounded responses due to likelihood maximization under joint probabilistic modeling.\n  We propose a stage-wise preference optimization framework for hallucination reduction through targeted multimodal data construction. Rather than directly optimizing on generic instruction-following data, our approach progressively constructs hallucination-focused preference pairs near known failure boundaries. The frame"},"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":"2605.16411","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-13T15:37:51Z","cross_cats_sorted":["cs.AI","cs.CL","cs.DB","cs.LG"],"title_canon_sha256":"b7e28bd1b5206b17ee1ac8ab61fc372c4e4819d8eaac63915679028baf21fbc4","abstract_canon_sha256":"fc595eb412fd72ea19196e81c9111a51f261fde757aba03c2e8f5061d4af12ef"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:20.940972Z","signature_b64":"NNo3GtFvOrAckW2Des8iiZKyyxVYSwiE+eXWcWIECwu+hqyS2RgDuP6NZEZqoTTLHmdqdJUXtl3WMSOfxOM9DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"45c99d0e679aa52dd3a8cc6a26761f5197e3835541385a743db5f3cecd9dc46b","last_reissued_at":"2026-05-20T00:02:20.940378Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:20.940378Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reducing Hallucination in Vision-Language Models via Stage-wise Preference Optimization under Distribution Shift","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.DB","cs.LG"],"primary_cat":"cs.CV","authors_text":"Qinwu Xu","submitted_at":"2026-05-13T15:37:51Z","abstract_excerpt":"Hallucination remains a fundamental challenge in vision-language models (VLMs), where autoregressive generation may produce linguistically plausible yet physically inconsistent or visually ungrounded responses due to likelihood maximization under joint probabilistic modeling.\n  We propose a stage-wise preference optimization framework for hallucination reduction through targeted multimodal data construction. Rather than directly optimizing on generic instruction-following data, our approach progressively constructs hallucination-focused preference pairs near known failure boundaries. The frame"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.16411","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/2605.16411/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-19T19:41:56.567066Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T19:34:36.595763Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"08ef8ed0310c601354ff83cbe032ac7f5a9069df35506ba64efef24bd8d1fbfb"},"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":"2605.16411","created_at":"2026-05-20T00:02:20.940458+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.16411v1","created_at":"2026-05-20T00:02:20.940458+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.16411","created_at":"2026-05-20T00:02:20.940458+00:00"},{"alias_kind":"pith_short_12","alias_value":"IXEZ2DTHTKSS","created_at":"2026-05-20T00:02:20.940458+00:00"},{"alias_kind":"pith_short_16","alias_value":"IXEZ2DTHTKSS3U5I","created_at":"2026-05-20T00:02:20.940458+00:00"},{"alias_kind":"pith_short_8","alias_value":"IXEZ2DTH","created_at":"2026-05-20T00:02:20.940458+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/IXEZ2DTHTKSS3U5IZRVCM5Q7KG","json":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG.json","graph_json":"https://pith.science/api/pith-number/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/graph.json","events_json":"https://pith.science/api/pith-number/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/events.json","paper":"https://pith.science/paper/IXEZ2DTH"},"agent_actions":{"view_html":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG","download_json":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG.json","view_paper":"https://pith.science/paper/IXEZ2DTH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.16411&json=true","fetch_graph":"https://pith.science/api/pith-number/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/graph.json","fetch_events":"https://pith.science/api/pith-number/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/action/storage_attestation","attest_author":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/action/author_attestation","sign_citation":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/action/citation_signature","submit_replication":"https://pith.science/pith/IXEZ2DTHTKSS3U5IZRVCM5Q7KG/action/replication_record"}},"created_at":"2026-05-20T00:02:20.940458+00:00","updated_at":"2026-05-20T00:02:20.940458+00:00"}