{"paper":{"title":"A Survey on Hallucination in Large Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Large vision-language models generate text that conflicts with input images, and this survey defines the problem while reviewing its symptoms, benchmarks, causes, and fixes.","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Dapeng Chen, Hanchao Liu, Ke Wang, Liping Hou, Rongjun Li, Wei Peng, Wenyuan Xue, Xiutian Zhao, Yifei Chen","submitted_at":"2024-02-01T00:33:21Z","abstract_excerpt":"Recent development of Large Vision-Language Models (LVLMs) has attracted growing attention within the AI landscape for its practical implementation potential. However, ``hallucination'', or more specifically, the misalignment between factual visual content and corresponding textual generation, poses a significant challenge of utilizing LVLMs. In this comprehensive survey, we dissect LVLM-related hallucinations in an attempt to establish an overview and facilitate future mitigation. Our scrutiny starts with a clarification of the concept of hallucinations in LVLMs, presenting a variety of hallu"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Hallucination, defined as the misalignment between factual visual content and corresponding textual generation, poses a significant challenge of utilizing LVLMs, and this survey establishes an overview to facilitate future mitigation.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That hallucinations can be consistently defined and isolated across diverse LVLM architectures and tasks without significant overlap with other error types such as reasoning failures.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"This survey reviews the definition, symptoms, evaluation benchmarks, root causes, and mitigation methods for hallucinations in large vision-language models.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Large vision-language models generate text that conflicts with input images, and this survey defines the problem while reviewing its symptoms, benchmarks, causes, and fixes.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"94927963cf9c0a8d68a781ebc484168fa21c4d54e9d5a0bf6136a10c04d26c83"},"source":{"id":"2402.00253","kind":"arxiv","version":2},"verdict":{"id":"bdbc8482-dc4c-481d-a809-04ade0131e21","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T22:05:38.517168Z","strongest_claim":"Hallucination, defined as the misalignment between factual visual content and corresponding textual generation, poses a significant challenge of utilizing LVLMs, and this survey establishes an overview to facilitate future mitigation.","one_line_summary":"This survey reviews the definition, symptoms, evaluation benchmarks, root causes, and mitigation methods for hallucinations in large vision-language models.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That hallucinations can be consistently defined and isolated across diverse LVLM architectures and tasks without significant overlap with other error types such as reasoning failures.","pith_extraction_headline":"Large vision-language models generate text that conflicts with input images, and this survey defines the problem while reviewing its symptoms, benchmarks, causes, and fixes."},"references":{"count":56,"sample":[{"doi":"","year":2022,"title":"Flamingo: a visual language model for few-shot learning","work_id":"994f306f-a295-42db-934f-7f15ed59e399","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond","work_id":"cbc2bb21-b6bb-46c0-80bf-107e195ffe10","ref_index":2,"cited_arxiv_id":"2308.12966","is_internal_anchor":true},{"doi":"","year":null,"title":"Position-enhanced visual instruction tuning for multimodal large language models","work_id":"70a11f00-50ba-409c-b9da-619abaed18ea","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"MiniGPT-v2: large language model as a unified interface for vision-language multi-task learning","work_id":"fb62cd1b-3991-40be-a987-3cfa5772b5b5","ref_index":4,"cited_arxiv_id":"2310.09478","is_internal_anchor":true},{"doi":"","year":null,"title":"InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks","work_id":"d9e035c7-9e23-4cc2-ad3e-be080fbbf2d9","ref_index":5,"cited_arxiv_id":"2312.14238","is_internal_anchor":true}],"resolved_work":56,"snapshot_sha256":"7ba873e9d79fbe63fdd9e6941ac8119a72ac6b9a9e74ca44ed6137c5a6884015","internal_anchors":21},"formal_canon":{"evidence_count":2,"snapshot_sha256":"bfa4b190e5b5a0a408b1642f593f179644c4f57adb77771a6c7811986ef8d003"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}