{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VRHORSS64RNOOX3LHONLN5L3VF","short_pith_number":"pith:VRHORSS6","schema_version":"1.0","canonical_sha256":"ac4ee8ca5ee45ae75f6b3b9ab6f57ba9735df29f670442ca28eeb327d1f9a57d","source":{"kind":"arxiv","id":"2502.12359","version":1},"attestation_state":"computed","paper":{"title":"LanP: Rethinking the Impact of Language Priors in Large Vision-Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Enyan Dai, Hongcheng Gao, Junjie Ao, Junjie Xu, Lei Feng, Minhua Lin, Qi Shi, Sike Fu, Suhang Wang, Xiang Zhang, Yilong Wang, Yuwei Niu, Zhifang Zhang, Zhiwei Zhang, Zongyu Wu","submitted_at":"2025-02-17T22:48:34Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) have shown impressive performance in various tasks. However, LVLMs suffer from hallucination, which hinders their adoption in the real world. Existing studies emphasized that the strong language priors of LVLMs can overpower visual information, causing hallucinations. However, the positive role of language priors is the key to a powerful LVLM. If the language priors are too weak, LVLMs will struggle to leverage rich parameter knowledge and instruction understanding abilities to complete tasks in challenging visual scenarios where visual information alone is"},"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":"2502.12359","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-02-17T22:48:34Z","cross_cats_sorted":[],"title_canon_sha256":"03337ed8b336c85f54345fa3356d8f63301f7c8d6b51a0dc220a8c3f41d2e7ac","abstract_canon_sha256":"87ff1ff8bf182369bc0809e96b12f39cd2fb684807f5830b1e47157c1c408041"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:54.362935Z","signature_b64":"xgQ9+Jewp9aiQs9xpbIavDUF6RWPnsmMSJtlhGwcAPyRbMIQnrDRYOFaXIrvuMlRBubuiLUPH8FdoWVMD25nBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ac4ee8ca5ee45ae75f6b3b9ab6f57ba9735df29f670442ca28eeb327d1f9a57d","last_reissued_at":"2026-07-05T10:15:54.362431Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:54.362431Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LanP: Rethinking the Impact of Language Priors in Large Vision-Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Enyan Dai, Hongcheng Gao, Junjie Ao, Junjie Xu, Lei Feng, Minhua Lin, Qi Shi, Sike Fu, Suhang Wang, Xiang Zhang, Yilong Wang, Yuwei Niu, Zhifang Zhang, Zhiwei Zhang, Zongyu Wu","submitted_at":"2025-02-17T22:48:34Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) have shown impressive performance in various tasks. However, LVLMs suffer from hallucination, which hinders their adoption in the real world. Existing studies emphasized that the strong language priors of LVLMs can overpower visual information, causing hallucinations. However, the positive role of language priors is the key to a powerful LVLM. If the language priors are too weak, LVLMs will struggle to leverage rich parameter knowledge and instruction understanding abilities to complete tasks in challenging visual scenarios where visual information alone is"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.12359","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/2502.12359/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":"2502.12359","created_at":"2026-07-05T10:15:54.362501+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.12359v1","created_at":"2026-07-05T10:15:54.362501+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.12359","created_at":"2026-07-05T10:15:54.362501+00:00"},{"alias_kind":"pith_short_12","alias_value":"VRHORSS64RNO","created_at":"2026-07-05T10:15:54.362501+00:00"},{"alias_kind":"pith_short_16","alias_value":"VRHORSS64RNOOX3L","created_at":"2026-07-05T10:15:54.362501+00:00"},{"alias_kind":"pith_short_8","alias_value":"VRHORSS6","created_at":"2026-07-05T10:15:54.362501+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06420","citing_title":"HoloCount: A Holistic Visual Counting Benchmark for MLLMs","ref_index":46,"is_internal_anchor":true},{"citing_arxiv_id":"2605.29602","citing_title":"CogniVerse: Revolutionizing Multi-Modal Retrieval-Augmented Generation with Cognitive Reflection and Geometric Reasoning","ref_index":111,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF","json":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF.json","graph_json":"https://pith.science/api/pith-number/VRHORSS64RNOOX3LHONLN5L3VF/graph.json","events_json":"https://pith.science/api/pith-number/VRHORSS64RNOOX3LHONLN5L3VF/events.json","paper":"https://pith.science/paper/VRHORSS6"},"agent_actions":{"view_html":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF","download_json":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF.json","view_paper":"https://pith.science/paper/VRHORSS6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.12359&json=true","fetch_graph":"https://pith.science/api/pith-number/VRHORSS64RNOOX3LHONLN5L3VF/graph.json","fetch_events":"https://pith.science/api/pith-number/VRHORSS64RNOOX3LHONLN5L3VF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF/action/storage_attestation","attest_author":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF/action/author_attestation","sign_citation":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF/action/citation_signature","submit_replication":"https://pith.science/pith/VRHORSS64RNOOX3LHONLN5L3VF/action/replication_record"}},"created_at":"2026-07-05T10:15:54.362501+00:00","updated_at":"2026-07-05T10:15:54.362501+00:00"}