{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:7RUOSTPSJ7MYU6S62ONS7DOLRE","short_pith_number":"pith:7RUOSTPS","schema_version":"1.0","canonical_sha256":"fc68e94df24fd98a7a5ed39b2f8dcb892810c3fecc966e4deb209cf0826c21af","source":{"kind":"arxiv","id":"2505.03420","version":2},"attestation_state":"computed","paper":{"title":"Mitigating Image Captioning Hallucinations in Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.MM","authors_text":"Chengcui Zhang, Fei Zhao, Runlin Zhang, Tianyang Wang, Xi Li","submitted_at":"2025-05-06T10:55:21Z","abstract_excerpt":"Hallucinations in vision-language models (VLMs) hinder reliability and real-world applicability, usually stemming from distribution shifts between pretraining data and test samples. Existing solutions, such as retraining or fine-tuning on additional data, demand significant computational resources and labor-intensive data collection, while ensemble-based methods incur additional costs by introducing auxiliary VLMs. To address these challenges, we propose a novel test-time adaptation framework using reinforcement learning to mitigate hallucinations during inference without retraining or any aux"},"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":"2505.03420","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.MM","submitted_at":"2025-05-06T10:55:21Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"493f781e308c566fc7c1a847bbf18f7dd889be204a2107c1be76d2ecf154211c","abstract_canon_sha256":"c29c1a75dfea0d1373b27c669a628f0fee63ae331ec8eac9502a8430aa43610d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:17:45.185749Z","signature_b64":"7u8bPrZt0BdbZxUKgG/IoAM8Jx5qwSkB/mulcNWJ2EV4zSRUCqbFfAXwrCmuyXtwn0qWMaB5d1lDREralkHhCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fc68e94df24fd98a7a5ed39b2f8dcb892810c3fecc966e4deb209cf0826c21af","last_reissued_at":"2026-07-05T11:17:45.185308Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:17:45.185308Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mitigating Image Captioning Hallucinations in Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.MM","authors_text":"Chengcui Zhang, Fei Zhao, Runlin Zhang, Tianyang Wang, Xi Li","submitted_at":"2025-05-06T10:55:21Z","abstract_excerpt":"Hallucinations in vision-language models (VLMs) hinder reliability and real-world applicability, usually stemming from distribution shifts between pretraining data and test samples. Existing solutions, such as retraining or fine-tuning on additional data, demand significant computational resources and labor-intensive data collection, while ensemble-based methods incur additional costs by introducing auxiliary VLMs. To address these challenges, we propose a novel test-time adaptation framework using reinforcement learning to mitigate hallucinations during inference without retraining or any aux"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.03420","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/2505.03420/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":"2505.03420","created_at":"2026-07-05T11:17:45.185377+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.03420v2","created_at":"2026-07-05T11:17:45.185377+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.03420","created_at":"2026-07-05T11:17:45.185377+00:00"},{"alias_kind":"pith_short_12","alias_value":"7RUOSTPSJ7MY","created_at":"2026-07-05T11:17:45.185377+00:00"},{"alias_kind":"pith_short_16","alias_value":"7RUOSTPSJ7MYU6S6","created_at":"2026-07-05T11:17:45.185377+00:00"},{"alias_kind":"pith_short_8","alias_value":"7RUOSTPS","created_at":"2026-07-05T11:17:45.185377+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.13156","citing_title":"Dual-Pathway Circuits of Object Hallucination in Vision-Language Models","ref_index":38,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE","json":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE.json","graph_json":"https://pith.science/api/pith-number/7RUOSTPSJ7MYU6S62ONS7DOLRE/graph.json","events_json":"https://pith.science/api/pith-number/7RUOSTPSJ7MYU6S62ONS7DOLRE/events.json","paper":"https://pith.science/paper/7RUOSTPS"},"agent_actions":{"view_html":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE","download_json":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE.json","view_paper":"https://pith.science/paper/7RUOSTPS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.03420&json=true","fetch_graph":"https://pith.science/api/pith-number/7RUOSTPSJ7MYU6S62ONS7DOLRE/graph.json","fetch_events":"https://pith.science/api/pith-number/7RUOSTPSJ7MYU6S62ONS7DOLRE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE/action/storage_attestation","attest_author":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE/action/author_attestation","sign_citation":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE/action/citation_signature","submit_replication":"https://pith.science/pith/7RUOSTPSJ7MYU6S62ONS7DOLRE/action/replication_record"}},"created_at":"2026-07-05T11:17:45.185377+00:00","updated_at":"2026-07-05T11:17:45.185377+00:00"}