{"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/mitigating-object-hallucination-via-data","title":"Data-augmented phrase-level alignment for mitigating object hallucination","arxiv_id":"2405.18654","date":"2024-05-28","proceeding":null,"authors":["Pritam Sarkar","Sayna Ebrahimi","Ali Etemad","Ahmad Beirami","Sercan Ö. Arik","Tomas Pfister"],"abstract":"Despite their significant advancements, Multimodal Large Language Models (MLLMs) often generate factually inaccurate information, referred to as hallucination. In this work, we address object hallucinations in MLLMs, where information is generated about an object not present in the input image. We introduce Data-augmented Phrase-level Alignment (DPA), a novel loss which can be applied to instruction-tuned off-the-shelf MLLMs to mitigate hallucinations, while preserving their general vision-language capabilities. To fine-tune MLLMs with DPA, we first generate a set of `hallucinated' and `correct' response pairs through generative data augmentation by selectively altering the ground-truth information of the correct responses at a phrase level. The DPA loss is then used to train MLLMs to reduce the likelihood of hallucinated phrases compared to the correct ones. Our thorough evaluation on various benchmarks confirms the effectiveness of DPA in mitigating hallucination while retaining the out-of-the-box performance of the MLLMs on general tasks. For instance, MLLMs finetuned with DPA, which we refer to as Hallucination Attenuated Language and Vision Assistant (HALVA), improve F1 by up to 13.4% on hallucination visual question-answering and reduce the hallucination rate by up to 4.2% on image description tasks.","url_abs":"https://arxiv.org/abs/2405.18654v2","url_pdf":"https://arxiv.org/pdf/2405.18654v2.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":[],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":"hallucination","task_name":"Hallucination"},{"task_slug":null,"task_name":"Image Description"},{"task_slug":"object","task_name":"Object"},{"task_slug":"object-hallucination","task_name":"Object Hallucination"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"}],"methods":[{"method_slug":"set","method_name":"SET"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2405.18654","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2405.18654"}},"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":"deterministic:regex_extraction","url":"https://github.com/pritamqu/HALVA","reach":{"status":"ok"}}],"summary":{"ran_draft_wrong":2,"ran":2,"ran_fixture":1,"unverified":5},"by_repo_kind":{"found_in_text":{"samples":10,"ran":5,"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":10,"samples":[{"code_sha256_prefix":"42a46570620cd9fa","entry":"get_chunk","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"eval_hall/model_amber_loader.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/eval_hall/model_amber_loader.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"42a46570620cd9fa"}},{"code_sha256_prefix":"4b9b7f2c9ec6c036","entry":"moving_average","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"llava/train/train_halva.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/llava/train/train_halva.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"4b9b7f2c9ec6c036"}},{"code_sha256_prefix":"9b4d5acd36382b39","entry":"parse_model_name_or_path","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"vila/model/builder_halva.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/vila/model/builder_halva.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9b4d5acd36382b39"}},{"code_sha256_prefix":"076c252c52cbb161","entry":"split_list","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"eval_hall/model_amber_loader.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/eval_hall/model_amber_loader.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"076c252c52cbb161"}},{"code_sha256_prefix":"10893c4608c08075","entry":"split_to_even_chunks","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"llava/train/halva_trainer.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/llava/train/halva_trainer.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"10893c4608c08075"}},{"code_sha256_prefix":"bb35e3ac741bb2c9","entry":"get_mm_adapter_state_maybe_zero_3","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"llava/train/halva_trainer.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/llava/train/halva_trainer.py","link_basis":"harvester_set","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":"bb35e3ac741bb2c9"}},{"code_sha256_prefix":"fa1225dfac92bc0d","entry":"get_peft_state_maybe_zero_3","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"llava/train/train_halva.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/llava/train/train_halva.py","link_basis":"harvester_set","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":"fa1225dfac92bc0d"}},{"code_sha256_prefix":"f4334934ee795269","entry":"is_transformers_greater_than","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"vila/train/halva_utils.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/vila/train/halva_utils.py","link_basis":"harvester_set","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":"f4334934ee795269"}},{"code_sha256_prefix":"735025744c1ab0cf","entry":"maybe_zero_3","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"llava/train/halva_trainer.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/llava/train/halva_trainer.py","link_basis":"harvester_set","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":"735025744c1ab0cf"}},{"code_sha256_prefix":"616ffbdc154ed2d8","entry":"maybe_zero_3","repo":"pritamqu/HALVA","repo_kind":"found_in_text","path":"llava/train/train_halva.py","file_url":"https://github.com/pritamqu/HALVA/blob/HEAD/llava/train/train_halva.py","link_basis":"harvester_set","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":"616ffbdc154ed2d8"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}