{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T19:47:59Z","timestamp":1771876079820,"version":"3.50.1"},"reference-count":58,"publisher":"Public Library of Science (PLoS)","issue":"2","license":[{"start":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T00:00:00Z","timestamp":1771804800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Ministry of Education of the Republic of Korea and the National Research Foundation of Korea","award":["NRF-2024S1A5C3A03046579"],"award-info":[{"award-number":["NRF-2024S1A5C3A03046579"]}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>Hallucinations are widely recognized as a significant drawback of large language models. Several attempts have been made to reduce the intensity of hallucinations. Among the various attempts, our research has been directed towards mitigating hallucinations caused by the co-occurrence statistics of pre-training corpora. We introduce Model-AGNostic countErfacTual synthesis and adaptive fine-tuning framework (MAGNET), a fine-tuning method that can mitigate the bias of co-occurrence statistics on large language models pre-training data when generating sentences. Our pipeline generates the counterfactual sample sentences and subject and object information for the counterfactual sample from the language model, and filters them to make sure they contain these three pieces of information before using them as fine-tuning data. Next, it utilizes both the generated counterfactual sample and the original sentence used to generate it as a training dataset. When our method is applied to GPT-Neo 2.7B model, it shows a 12% improvement in the Factual Knowledge Probing experiment, and there is a correlation analysis that can mitigate the bias on the pre-training data. In the TruthfulQA experiment, when fine-tuning the GPT-Neo 125M model on the LAMA-TREx dataset, applying our method showed 2.27% better performance than not applying it.<\/jats:p>","DOI":"10.1371\/journal.pone.0340812","type":"journal-article","created":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T18:41:20Z","timestamp":1771872080000},"page":"e0340812","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["MAGNET: Counterfactual samples synthesizing for mitigating hallucination in large language models"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0009-0006-3260-2307","authenticated-orcid":true,"given":"Byeong Su","family":"Kim","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Beomsoo","family":"Kim","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3911-5935","authenticated-orcid":true,"given":"Beakcheol","family":"Jang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2026,2,23]]},"reference":[{"key":"pone.0340812.ref001","article-title":"Llama 2: open foundation and fine-tuned chat models","author":"H Touvron","year":"2023","journal-title":"arXiv preprint"},{"key":"pone.0340812.ref002","article-title":"A survey of large language models","author":"WX Zhao","year":"2023","journal-title":"arXiv preprint"},{"key":"pone.0340812.ref003","article-title":"Entailment as few-shot learner","author":"S Wang","year":"2021","journal-title":"arXiv preprint"},{"key":"pone.0340812.ref004","doi-asserted-by":"crossref","unstructured":"Nair I, Somasundaram S, Saxena A, Goswami K. Drilling down into the discourse structure with LLMs for long document question answering. In: Findings of the Association for Computational Linguistics: EMNLP 2023 . 2023. p. 9548\u201366.","DOI":"10.18653\/v1\/2023.findings-emnlp.972"},{"key":"pone.0340812.ref005","article-title":"A survey of chain of thought reasoning: advances, frontiers and future","author":"Z Chu","year":"2023","journal-title":"arXiv preprint"},{"key":"pone.0340812.ref006","doi-asserted-by":"crossref","unstructured":"Wang L, Lyu C, Ji T, Zhang Z, Yu D, Shi S, et al. Document-level machine translation with large language models. In: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2023. p. 16646\u201361.","DOI":"10.18653\/v1\/2023.emnlp-main.1036"},{"key":"pone.0340812.ref007","unstructured":"Welleck S, Kulikov I, Roller S, Dinan E, Cho K, Weston J. Neural text generation with unlikelihood training. In: 8th International Conference on Learning Representations (ICLR); 2020."},{"key":"pone.0340812.ref008","doi-asserted-by":"crossref","unstructured":"Raunak V, Menezes A, Junczys-Dowmunt M. The curious case of hallucinations in neural machine translation. In: Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL). 2021. p. 1172\u201383.","DOI":"10.18653\/v1\/2021.naacl-main.92"},{"key":"pone.0340812.ref009","unstructured":"Singla S, Feizi S. Salient imagenet: How to discover spurious features in deep learning? In: International Conference on Learning Representations (ICLR); 2022."},{"key":"pone.0340812.ref010","doi-asserted-by":"crossref","unstructured":"Kang C, Choi J. Impact of co-occurrence on factual knowledge of large language models. In: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2023. p. 2795\u2013813.","DOI":"10.18653\/v1\/2023.findings-emnlp.518"},{"key":"pone.0340812.ref011","doi-asserted-by":"crossref","unstructured":"Niu Y, Tang K, Zhang H, Lu Z, Hua XS, Wen JR. Counterfactual VQA: a cause-effect look at language bias. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2021. p. 12700\u201310.","DOI":"10.1109\/CVPR46437.2021.01251"},{"key":"pone.0340812.ref012","doi-asserted-by":"crossref","unstructured":"Hong P, Bhardwaj R, Majumdar N, Aditya S, Poria S. ReMask: a robust information-masking approach for domain counterfactual generation. In: Proceedings of the 2023 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL). 2024.","DOI":"10.18653\/v1\/2023.findings-acl.231"},{"key":"pone.0340812.ref013","doi-asserted-by":"crossref","unstructured":"Calderon N, Ben-David E, Feder A, Reichart R. Docogen: domain counterfactual generation for low resource domain adaptation. In: Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022. p. 7727\u201346.","DOI":"10.18653\/v1\/2022.acl-long.533"},{"key":"pone.0340812.ref014","unstructured":"Ou J, Zhang J, Feng Y, Zhou J. Counterfactual data augmentation via perspective transition for open-domain dialogues. In: Proceedings of the 29th International Conference on Computational Linguistics (COLING). 2022. p. 3254\u201364."},{"key":"pone.0340812.ref015","doi-asserted-by":"crossref","unstructured":"Paranjape B, Lamm M, Tenney I. Retrieval-guided counterfactual generation for QA. In: Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022. p. 1670\u201386.","DOI":"10.18653\/v1\/2022.acl-long.117"},{"issue":"12","key":"pone.0340812.ref016","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3571730","article-title":"Survey of hallucination in natural language generation","volume":"55","author":"Z Ji","year":"2023","journal-title":"ACM Computing Surveys."},{"key":"pone.0340812.ref017","doi-asserted-by":"crossref","unstructured":"Rajaee S, Yaghoobzadeh Y, Pilehvar MT. Looking at the overlooked: an analysis on the word-overlap bias in natural language inference. In: Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL). 2022. p. 4247\u201363.","DOI":"10.18653\/v1\/2022.emnlp-main.725"},{"key":"pone.0340812.ref018","doi-asserted-by":"crossref","unstructured":"Kassner N, Sch\u00fctze H. Negated and misprimed probes for pretrained language models: Birds can talk, but cannot fly. In: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL). 2020. p. 7811\u20138.","DOI":"10.18653\/v1\/2020.acl-main.698"},{"key":"pone.0340812.ref019","doi-asserted-by":"crossref","unstructured":"Poerner N, Waltinger U, Sch\u00fctze H. E-BERT: Efficient-yet-effective entity embeddings for BERT. In: Findings of the Association for Computational Linguistics: EMNLP 2020 . 2020. p. 803\u201318.","DOI":"10.18653\/v1\/2020.findings-emnlp.71"},{"issue":"4940","key":"pone.0340812.ref020","doi-asserted-by":"crossref","first-page":"301","DOI":"10.1126\/science.2296719","article-title":"Priming and human memory systems","volume":"247","author":"E Tulving","year":"1990","journal-title":"Science."},{"key":"pone.0340812.ref021","doi-asserted-by":"crossref","unstructured":"Durmus E, Ladhak F, Hashimoto T. Spurious correlations in reference-free evaluation of text generation.In: Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022. p. 1443\u201354.","DOI":"10.18653\/v1\/2022.acl-long.102"},{"key":"pone.0340812.ref022","doi-asserted-by":"crossref","unstructured":"Khani F, Liang P. Removing spurious features can hurt accuracy and affect groups disproportionately. In: Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency; 2021. p. 196\u2013205.","DOI":"10.1145\/3442188.3445883"},{"key":"pone.0340812.ref023","doi-asserted-by":"crossref","unstructured":"Chen Z, Gao Q, Bosselut A, Sabharwal A, Richardson K. Disco: Distilling counterfactuals with large language models. In: Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023. p. 5514\u201328.","DOI":"10.18653\/v1\/2023.acl-long.302"},{"key":"pone.0340812.ref024","unstructured":"Kaushik D, Hovy E, Lipton ZC. Learning the difference that makes a difference with counterfactually-augmented data. In: International Conference on Learning Representations (ICLR); 2020."},{"key":"pone.0340812.ref025","unstructured":"Bauer A, Hoedoro N, Schneider A. Rule-based Approach to Text Generation in Natural Language-Automated Text Markup Language (ATML3). In: Challenge+ DC@ RuleML; 2015."},{"issue":"3","key":"pone.0340812.ref026","doi-asserted-by":"crossref","first-page":"171","DOI":"10.1016\/S0167-6393(98)00018-1","article-title":"A study of n-gram and decision tree letter language modeling methods","volume":"24","author":"G Potamianos","year":"1998","journal-title":"Speech Communication."},{"issue":"13","key":"pone.0340812.ref027","doi-asserted-by":"crossref","first-page":"10575","DOI":"10.1007\/s00521-022-07184-7","article-title":"Rule-based adversarial sample generation for text classification","volume":"34","author":"N Zhou","year":"2022","journal-title":"Neural Computing and Applications."},{"key":"pone.0340812.ref028","doi-asserted-by":"crossref","unstructured":"Wu T, Ribeiro MT, Heer J, Weld DS. Polyjuice: Generating counterfactuals for explaining, evaluating, and improving models. In: Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics (ACL). 2021. p. 6707\u201323.","DOI":"10.18653\/v1\/2021.acl-long.523"},{"issue":"8","key":"pone.0340812.ref029","first-page":"9","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"A Radford","year":"2019","journal-title":"OpenAI Blog."},{"key":"pone.0340812.ref030","article-title":"Flexible text generation for counterfactual fairness probing","author":"Z Fryer","year":"2022","journal-title":"arXiv preprint"},{"key":"pone.0340812.ref031","article-title":"LaMDA: language models for dialog applications","author":"R Thoppilan","year":"2022","journal-title":"arXiv preprint"},{"key":"pone.0340812.ref032","first-page":"1877","article-title":"Language models are few-shot learners. In:","author":"T Brown","year":"2020","journal-title":"Advances in Neural Information Processing Systems (NeurIPS)."},{"key":"pone.0340812.ref033","first-page":"1","article-title":"GPT understands, too","volume":"4","author":"X Liu","year":"2023","journal-title":"AI Open."},{"key":"pone.0340812.ref034","doi-asserted-by":"crossref","unstructured":"Lester B, Al-Rfou R, Constant N. The power of scale for parameter-efficient prompt tuning. In: Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2021. p. 3045\u201359.","DOI":"10.18653\/v1\/2021.emnlp-main.243"},{"key":"pone.0340812.ref035","doi-asserted-by":"crossref","unstructured":"Li XL, Liang P. Prefix-tuning: Optimizing continuous prompts for generation. In: Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics (ACL). 2021. p. 4582\u201397.","DOI":"10.18653\/v1\/2021.acl-long.353"},{"key":"pone.0340812.ref036","doi-asserted-by":"crossref","unstructured":"Qin G, Eisner J. Learning how to ask: querying LMs with mixtures of soft prompts. In: Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL). 2021. p. 5203\u201319.","DOI":"10.18653\/v1\/2021.naacl-main.410"},{"key":"pone.0340812.ref037","doi-asserted-by":"crossref","unstructured":"Liu X, Ji K, Fu Y, Tam W, Du Z, Yang Z, et al. P-tuning: Prompt tuning can be comparable to fine-tuning across scales and tasks. In: Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers). 2022. p. 61\u20138.","DOI":"10.18653\/v1\/2022.acl-short.8"},{"issue":"9","key":"pone.0340812.ref038","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3560815","article-title":"Pre-train, prompt, and predict: a systematic survey of prompting methods in natural language processing","volume":"55","author":"P Liu","year":"2023","journal-title":"ACM Computing Surveys."},{"key":"pone.0340812.ref039","doi-asserted-by":"crossref","unstructured":"Bouraoui Z, Camacho-Collados J, Schockaert S. Inducing relational knowledge from BERT. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 34; 2020. p. 7456-63.","DOI":"10.1609\/aaai.v34i05.6242"},{"key":"pone.0340812.ref040","doi-asserted-by":"crossref","first-page":"423","DOI":"10.1162\/tacl_a_00324","article-title":"How can we know what language models know?","volume":"8","author":"Z Jiang","year":"2020","journal-title":"Transactions of the Association for Computational Linguistics."},{"key":"pone.0340812.ref041","doi-asserted-by":"crossref","unstructured":"Haviv A, Berant J, Globerson A. BERTese: learning to speak to BERT. In: Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL). 2021. p. 3618\u201323.","DOI":"10.18653\/v1\/2021.eacl-main.316"},{"key":"pone.0340812.ref042","doi-asserted-by":"crossref","unstructured":"Shin T, Razeghi Y, Logan IV RL, Wallace E, Singh S. Autoprompt: Eliciting knowledge from language models with automatically generated prompts. In: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2020. p. 4222\u201335.","DOI":"10.18653\/v1\/2020.emnlp-main.346"},{"key":"pone.0340812.ref043","unstructured":"Fichtel L, Kalo JC, Balke WT. Prompt tuning or fine-tuning: investigating relational knowledge in pre-trained language models. In: 3rd Conference on Automated Knowledge Base Construction (AKBC); 2021."},{"key":"pone.0340812.ref044","unstructured":"Sanh V, Webson A, Raffel C, Bach SH, Sutawika L, Alyafeai Z, et al. Multitask prompted training enables zero-shot task generalization. In: International Conference on Learning Representations (ICLR); 2022."},{"key":"pone.0340812.ref045","doi-asserted-by":"crossref","unstructured":"Lai X, Tian Z, Chen Y, Li Y, Yuan Y, Liu S, et al. Lisa: Reasoning segmentation via large language model. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2024.","DOI":"10.1109\/CVPR52733.2024.00915"},{"key":"pone.0340812.ref046","doi-asserted-by":"crossref","unstructured":"Liu J, Zhang Z, Xiao J, Jin Z, Zhang X, Ma Y, et al. Large Language Model Locally Fine-tuning (LLMLF) on Chinese Medical Imaging Reports. In: Proceedings of the 2023 6th International Conference on Big Data Technologies. 2023. p. 273\u20139.","DOI":"10.1145\/3627377.3627445"},{"key":"pone.0340812.ref047","article-title":"GPT-Neo: Large scale autoregressive language modeling with mesh-tensorflow","author":"S Black","year":"2021","journal-title":"Zenodo."},{"key":"pone.0340812.ref048","article-title":"The Pile: an 800GB dataset of diverse text for language modeling","author":"L Gao","year":"2020","journal-title":"arXiv preprint"},{"key":"pone.0340812.ref049","doi-asserted-by":"crossref","unstructured":"Liu Y, Wan Y, He L, Peng H, Philip SY. KG-BART: knowledge graph-augmented BART for generative commonsense reasoning. In: Proceedings of the AAAI Conference on Artificial Intelligence. 2021. p. 6418\u201325.","DOI":"10.1609\/aaai.v35i7.16796"},{"key":"pone.0340812.ref050","doi-asserted-by":"crossref","unstructured":"Bauer L, Wang Y, Bansal M. Commonsense for generative multi-hop question answering tasks. In: Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing (EMNLP); 2018. p. 4220\u201330.","DOI":"10.18653\/v1\/D18-1454"},{"key":"pone.0340812.ref051","doi-asserted-by":"crossref","unstructured":"Cheng L, Wu D, Bing L, Zhang Y, Jie Z, Lu W, et al. Ent-desc: Entity description generation by exploring knowledge graph. In: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2020. p. 1194\u2013205.","DOI":"10.18653\/v1\/2020.emnlp-main.90"},{"key":"pone.0340812.ref052","unstructured":"Kim B, Ahn J, Kim G. Sequential latent knowledge selection for knowledge-grounded dialogue. In: International Conference on Learning Representations (ICLR); 2020."},{"key":"pone.0340812.ref053","doi-asserted-by":"crossref","unstructured":"Krishna K, Roy A, Iyyer M. Hurdles to progress in long-form question answering. In: Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL); 2021. p. 4940\u201357.","DOI":"10.18653\/v1\/2021.naacl-main.393"},{"key":"pone.0340812.ref054","doi-asserted-by":"crossref","unstructured":"Kim J, Choi S, Amplayo RK, Hwang Sw. Retrieval-augmented controllable review generation. In: Proceedings of the 28th International Conference on Computational Linguistics (COLING). 2020. p. 2284\u201395.","DOI":"10.18653\/v1\/2020.coling-main.207"},{"key":"pone.0340812.ref055","doi-asserted-by":"crossref","unstructured":"Ghazvininejad M, Brockett C, Chang MW, Dolan B, Gao J, Yih Wt, et al. A knowledge-grounded neural conversation model. In: Proceedings of the AAAI Conference on Artificial Intelligence. 2018.","DOI":"10.1609\/aaai.v32i1.11977"},{"key":"pone.0340812.ref056","doi-asserted-by":"crossref","unstructured":"Veseli B, Razniewski S, Kalo JC, Weikum G. Evaluating the knowledge base completion potential of GPT. In: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2023. p. 11099\u2013112.","DOI":"10.18653\/v1\/2023.findings-emnlp.426"},{"key":"pone.0340812.ref057","doi-asserted-by":"crossref","first-page":"400","DOI":"10.1214\/aoms\/1177729586","article-title":"A stochastic approximation method","author":"H Robbins","year":"1951","journal-title":"The Annals of Mathematical Statistics."},{"key":"pone.0340812.ref058","doi-asserted-by":"crossref","unstructured":"Wolf T, Debut L, Sanh V, Chaumond J, Delangue C, Moi A, et al. Transformers: state-of-the-art natural language processing. In: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations. 2020. p. 38\u201345.","DOI":"10.18653\/v1\/2020.emnlp-demos.6"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0340812","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T18:41:37Z","timestamp":1771872097000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0340812"}},"subtitle":[],"editor":[{"given":"Sonia","family":"Vasconcelos","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2026,2,23]]},"references-count":58,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2026,2,23]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0340812","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,23]]}}}