{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/generate-but-verify-reducing-hallucination-in","title":"Generate, but Verify: Reducing Hallucination in Vision-Language Models with Retrospective Resampling","arxiv_id":"2504.13169","date":"2025-04-17","proceeding":null,"authors":["Tsung-Han Wu","HeeKyung Lee","Jiaxin Ge","Joseph E. Gonzalez","Trevor Darrell","David M. Chan"],"abstract":"Vision-Language Models (VLMs) excel at visual understanding but often suffer from visual hallucinations, where they generate descriptions of nonexistent objects, actions, or concepts, posing significant risks in safety-critical applications. Existing hallucination mitigation methods typically follow one of two paradigms: generation adjustment, which modifies decoding behavior to align text with visual inputs, and post-hoc verification, where external models assess and correct outputs. While effective, generation adjustment methods often rely on heuristics and lack correction mechanisms, while post-hoc verification is complicated, typically requiring multiple models and tending to reject outputs rather than refine them. In this work, we introduce REVERSE, a unified framework that integrates hallucination-aware training with on-the-fly self-verification. By leveraging a new hallucination-verification dataset containing over 1.3M semi-synthetic samples, along with a novel inference-time retrospective resampling technique, our approach enables VLMs to both detect hallucinations during generation and dynamically revise those hallucinations. Our evaluations show that REVERSE achieves state-of-the-art hallucination reduction, outperforming the best existing methods by up to 12% on CHAIR-MSCOCO and 28% on HaloQuest. Our dataset, model, and code are available at: https://reverse-vlm.github.io.","url_abs":"https://arxiv.org/abs/2504.13169v1","url_pdf":"https://arxiv.org/pdf/2504.13169v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"generate-but-verify-reducing-hallucination-in","repo_url":"https://github.com/tsunghan-wu/reverse_vlm","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"hallucination","task_name":"Hallucination"}],"methods":[{"method_slug":"align","method_name":"ALIGN"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2504.13169","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2504.13169"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tsunghan-wu/reverse_vlm","reach":null}],"summary":{"ran_draft_wrong":6,"ran_violates":1,"ran_fixture":3,"unverified":1},"by_repo_kind":{"official":{"samples":9,"ran":8,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":11,"samples":[{"code_sha256_prefix":"3a19bf415091befc","entry":"back_track_last_sentence","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"llava/eval/inference_utils.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/llava/eval/inference_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"3a19bf415091befc"}},{"code_sha256_prefix":"194dfaa5e36a26ea","entry":"check_special_token_mismatch","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"llava/eval/inference_utils.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/llava/eval/inference_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"194dfaa5e36a26ea"}},{"code_sha256_prefix":"25585704d77cafb4","entry":"combine_coco_captions","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"25585704d77cafb4"}},{"code_sha256_prefix":"66415dc1dd50f555","entry":"combine_coco_instances","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"scripts/eval/eval_chair.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/scripts/eval/eval_chair.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"66415dc1dd50f555"}},{"code_sha256_prefix":"527ddfe8465dda89","entry":"de_hallucination_postprocessing","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"llava/eval/inference_utils.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/llava/eval/inference_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"527ddfe8465dda89"}},{"code_sha256_prefix":"f13b7a81d4dfe183","entry":"do_inference","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"llava/eval/inference_utils.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/llava/eval/inference_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f13b7a81d4dfe183"}},{"code_sha256_prefix":"f7756cda9b536a1b","entry":"early_exit","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"llava/eval/inference_utils.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/llava/eval/inference_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f7756cda9b536a1b"}},{"code_sha256_prefix":"897c16b1c81103f4","entry":"load_generated_captions","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"scripts/eval/eval_chair.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/scripts/eval/eval_chair.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"897c16b1c81103f4"}},{"code_sha256_prefix":"ee5e60b564c5eff5","entry":"render_output","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"llava/eval/inference_utils.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/llava/eval/inference_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ee5e60b564c5eff5"}},{"code_sha256_prefix":"55c32993da87759b","entry":"unpad_image","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"55c32993da87759b"}},{"code_sha256_prefix":"f8f2edbc52243524","entry":"retrospective_resampling_generate","repo":"tsunghan-wu/reverse_vlm","repo_kind":"official","path":"llava/eval/inference_utils.py","file_url":"https://github.com/tsunghan-wu/reverse_vlm/blob/HEAD/llava/eval/inference_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f8f2edbc52243524"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}