Survey of Hallucination in Natural Language Generation
Open accessThe DOI matches its Crossref record; 355 of the 360 references it cites were found in a registry or an index, and 4 are not checked yet; 224 of the 256 sources its claims cite were checked for support.
Abstract
Natural Language Generation (NLG) has improved exponentially in recent years thanks to the development of sequence-to-sequence deep learning technologies such as Transformer-based language models. This advancement has led to more fluent and coherent NLG, leading to improved development in downstream tasks such as abstractive summarization, dialogue generation, and data-to-text generation. However, it is also apparent that deep learning based generation is prone to hallucinate unintended text, which degrades the system performance and fails to meet user expectations in many real-world scenarios. To address this issue, many studies have been presented in measuring and mitigating hallucinated texts, but these have never been reviewed in a comprehensive manner before. In this survey, we thus provide a broad overview of the research progress and challenges in the hallucination problem in NLG. The survey is organized into two parts: (1) a general overview of metrics, mitigation methods, and future directions, and (2) an overview of task-specific research progress on hallucinations in the following downstream tasks, namely abstractive summarization, dialogue generation, generative question answering, data-to-text generation, and machine translation. This survey serves to facilitate collaborative efforts among researchers in tackling the challenge of hallucinated texts in NLG.
Verified 2026-11-19 · 4 references not checked yet
Overview
CoChat Summary
TL;DR
This survey provides a broad overview of hallucination in natural language generation, covering definitions, contributors, metrics, mitigation methods, and task-specific progress across summarization, dialogue, GQA, data-to-text, machine translation, vision-language generation, and LLMs. It distinguishes intrinsic from extrinsic hallucination and argues that hallucination must be treated differently across tasks and sources.
Key findings
- The survey categorizes hallucination into intrinsic hallucination, where generated output contradicts the source, and extrinsic hallucination, where output cannot be verified from the source; extrinsic hallucination is not always erroneous because it may come from factually correct external information.
- It reports prevalence examples from cited work: 25% of summaries generated by state-of-the-art abstractive summarization systems have hallucinated content; 62% of first sentences in WIKIBIO have additional information not stated in the corresponding infobox; and 74% of samples in the QMSum dataset consist of inconsistent facts.
- It reviews hallucination metrics as statistical, model-based (IE-based, QA-based, NLI-based, faithfulness classification, LM-based), and human evaluation, and mitigation methods as data-related and modeling/inference methods including architecture changes, planning/sketching, RL, multi-task learning, controllable generation, and post-processing.
- For LLMs, it notes that RLHF reduced the hallucination rate from 41% to 21% in closed-domain tasks for InstructGPT, and discusses reference-dependent and reference-free hallucination metrics plus mitigation via data refinement, RAG, safety fine-tuning, model editing, decoding methods, chain-of-thought variants, post-processing, and ensembling.
Methods
- Literature survey design: the paper reviews existing research on hallucination in NLG rather than conducting new experiments or primary data collection.
- Coverage: general NLG hallucination topics (definitions, contributors, metrics, mitigation, future directions) plus task-specific sections on abstractive summarization, dialogue generation, generative question answering, data-to-text generation, neural machine translation, vision-language generation, and large language models.
- Analysis: synthesizes and compares definitions, metrics, mitigation approaches, and future directions across tasks, with examples in tables and figures; the LLM section was updated in Jan 2024.
Limitations
- The paper states that GQA has no standard definition or categorization of hallucination and no automatic metric specifically for hallucination in GQA; it also notes that open-domain dialogue lacks a standard hallucination metric and that VL hallucination research is still in its early stage.
- The paper states that extrinsic hallucination has been somewhat overlooked and is more challenging to reduce, and that most existing mitigation methods do not distinguish between intrinsic and extrinsic hallucination.
- Evident from the text: as a survey, it does not provide new empirical validation of metrics or mitigation methods, and its LLM coverage is explicitly updated only through Jan 2024, so later work is not included.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures
Verification
Claims the paper makes
124 Valid · 26 Not valid · 19 Not checked · 135 Own results and interpretations
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“Furthermore, perturbation, where the insertions of most common tokens are placed at the beginning of the sentence, seems to be the most successful in hallucination mitigation.”
What the sources say
- Not checked — no full text of the sourcecites [133] [133] Katherine Lee, Orhan Firat, Ashish Agarwal, Clara Fannjiang, and David Sussillo. 2019. Hallucinations in Neural Machine Translation. ICLR (2019). -
- Disputedcites [215] [215] Vikas Raunak, Arul Menezes, and Marcin Junczys-Dowmunt. 2021. The Curious Case of Hallucinations in Neural Machine Translation. (4 2021), 1172-1183. https://arxiv.org/abs/2104.06683v1 - sourceExperimental — accuracy not yet measured
In the passage describing the perturbation procedure (page unavailable), the paper says that inserting common tokens at the beginning is used to generate hallucinations, not mitigate them. Its prevention section instead discusses approaches such as robust learning and corpus-level filtering. (LLM, conf=0.92)
Checked by a reasoning model — may be wrong.
hedgedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“While there are works on the visual extension of faithfulness-oriented RLHF [303], these essentially involve labeling more data, which is expensive and difficult to scale.”
What the sources say
- Disputedcites [303] [303] Tianyu Yu, Yuan Yao, Haoye Zhang, Taiwen He, Yifeng Han, Ganqu Cui, Jinyi Hu, Zhiyuan Liu, Hai-Tao Zheng, Maosong Sun, and Tat-Seng Chua. 2023. RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback. CoRR abs/2312.00849 (2023). https://doi.org/10.48550/ARXIV.2312.00849 arXiv:2312.00849 - sourceExperimental — accuracy not yet measured
RLHF-V does use human annotation for visual-model alignment, but it makes the contrary point about needing more labeled data. The Fine-grained Correctional Human Preference Collection section says its corrections improve data efficiency, and Section 4.3 reports comparable hallucination rates using only 200 preference examples versus an order of magnitude more ranking-labeled examples. Our copy does not establish a separate claim about annotation cost or scalability. (LLM, conf=0.88)
Checked by a reasoning model — may be wrong.
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
Every source the paper cites for this claim was checked and supports it.
In the paper
“Nevertheless, as illustrated by Falke et al. [66], off-the-shelf NLI models tend to transfer poorly to the abstractive summarization task.”
What the sources say
- Supportedcites [66] [66] Tobias Falke, Leonardo FR Ribeiro, Prasetya Ajie Utama, Ido Dagan, and Iryna Gurevych. 2019. Ranking Generated Summaries by Correctness: An Interesting but Challenging Application for Natural Language Inference. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics . 2214-2220. - source
The Introduction states that models trained on NLI datasets transfer poorly to the paper’s downstream task of ranking generated summaries by correctness. The Conclusions likewise say that out-of-the-box NLI models do not perform well on that task. (LLM, conf=0.94)
- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Such wrong association results in generation with facts mixed up between two similar entities [60, 233].”
What the sources say
- Partly supportedcites [60] [60] Nouha Dziri, Andrea Madotto, Osmar Zaiane, and Avishek Joey Bose. 2021. Neural Path Hunter: Reducing Hallucination in Dialogue Systems via Path Grounding. EMNLP (2021). - sourceExperimental — accuracy not yet measured
The GPT2-KG passage describes an incorrect association between entities, and G Error Analysis gives an example of a generated response assigning a book to the wrong author. These support the claim that facts can be mixed up, but the paper does not say the entities involved are similar. The supplied passages do not provide page numbers. (LLM, conf=0.83)
- Supportedcites [233] [233] Kurt Shuster, Spencer Poff, Moya Chen, Douwe Kiela, and Jason Weston. 2021. Retrieval Augmentation Reduces Hallucination in Conversation. EMNLP (2021). - source
The Introduction states that language models often mix up facts between two similar entities when generating factually incorrect statements. This directly supports the cited description of the resulting error. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“As a result, the target reference may contain information that cannot be supported by the source [198, 268].”
What the sources say
- Supportedcites [198] [198] Ankur Parikh, Xuezhi Wang, Sebastian Gehrmann, Manaal Faruqui, Bhuwan Dhingra, Diyi Yang, and Dipanjan Das. 2020. ToTTo: A Controlled Table-To-Text Generation Dataset. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) . 1173-1186. - source
The Dataset section says that reference targets selected from web sentences may contain information that cannot be inferred from the source. The Data Annotation Process section also describes deleting phrases unsupported by table cells. (LLM, conf=0.93)
- Supportedcites [268] [268] Hongmin Wang. 2019. Revisiting Challenges in Data-to-Text Generation with Fact Grounding. In Proceedings of the 12th International Conference on Natural Language Generation . 311-322. - source
The Introduction reports that about 40% of human-written game-summary content cannot be directly mapped to the input boxscore records and draws on information beyond those inputs. Section 2.1 likewise describes reference content not licensed by the boxscore table. This supports the claim that a target reference may contain information the source does not support. (LLM, conf=0.93)
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
117 more claims · 74 with their cited sources checked
Create a free account to see all 304 claims and to verify the 2 references still open.
- Borderline: one of the tests for a substantial claim is unclear. 1
- It may state two separate claims as one. 90
- Probably not a substantial claim. 142
- The quote only partly states this claim. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Hallucination can lead to serious consequences such as incorrect or misleading code, security risks, and ethical issues [186].”
From A survey on large language model based autonomous agents (2024) · cites this paper as [186]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“citation fabrication constitutes a form of extrinsic hallucination in the absence of retrieval augmentation, since the model generates references not grounded in any provided source material [4].”
From How LLMs Cite and Why It Matters: A Cross-Model Audit of Reference Fabrication in AI-Assisted Academic Writing and Methods to Detect Phantom Citations (2026) · cites this paper as [4]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Recent taxonomic work distinguishes at least seven cases [15,18].”
From Hallucination as epistemic pollution: epistemic injustice and credibility in futures-oriented science education (2026) · cites this paper as [18]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“LLMs struggle to generate accurate and up-to-date responses on current topics, due to outdated knowledge, lack of domain-specific expertise, or hallucination13–17.”
From Leveraging long context in retrieval augmented language models for medical question answering (2025) · cites this paper as [16]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Fine-tuning can enrich LLMs with domain knowledge, enhance their specific ability, improve the fairness and reliability of the outputs, and prevent certain damage caused by hallucination [26].”
From FederatedScope-LLM: A Comprehensive Package for Fine-tuning Large Language Models in Federated Learning (2024) · cites this paper as [26]
Experimental — accuracy not yet measured
41 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Extrinsic Object Hallucination
“CHAIR (Caption Hallucination Assessment with Image Relevance) metric, which calculates what proportion of object words generated are actually in the image according to the ground truth captions.”
§ Extrinsic Object Hallucination
“CHAIR 𝑖 measures per-instance object hallucination, i.e. what fraction of object instances in each generated caption are hallucinated.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 7.1 Hallucination Definition in Abstractive Summarization
“For instance, in Table 1, given the input article shown in the caption, an example of intrinsic hallucination is ' The Ebola vaccine was rejected by the FDA in 2019, '”
§ 2.1 Categorization
“in the abstractive summarization task from Table 1, the generated summary ' The first Ebola vaccine was approved in 2021 ' contradicts the source content”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Survey of Hallucination in Natural Language Generation
“The survey is organized into two parts: (1) a general overview of metrics, mitigation methods, and future directions; (2) an overview of task-specific research progress on hallucinations in the following downstream tasks, namely abstractive summarization, dialogue generation, generative question answering, data-to-text generation, machine translation, and visual-language generation; and (3) hallucinations in large language models (LLMs)”
§ 1 INTRODUCTION
“The third part discusses this phenomenon in LLMs in Section 13.”
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325 · cited 3×
“Recently, the rapid development of NLG has captured the imagination of many thanks to the advances in deep learning technologies, especially Transformer [189]-based models like BERT [29], BART [100], GPT-2 [149], and GPT-3 [16].”
BackgroundAn Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. (2020) · arXiv · doi:10.48550/arxiv.2010.11929 · Cited by 47,472
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Sequence to Sequence Learning with Neural Networks
Sutskever, I., Vinyals, O., & Le, Q. V. (2014) · arXiv · doi:10.48550/arxiv.1409.3215 · Cited by 17,766
Advances in Neural Information Processing Systems 29
Teymur, O., Zygalakis, K., & Calderhead, B. (2016) · Cited by 13,887
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019) · arXiv · doi:10.48550/arxiv.1910.10683 · Cited by 10,929 · cited 1×
“This work investigates the performances of GPT-3 [13], GPT-Neo/J [141], GPT-2 [105] and a T5-based model [106].”
BackgroundAdvances in Neural Information Processing Systems 14: Proceedings of the 2001 Conference
(2002) · The MIT Press eBooks · doi:10.7551/mitpress/1120.001.0001 · Cited by 8,890
Lewis, M., Liu, Y., Goyal, N., Ghazvininejad, M., Mohamed, A., Levy, O., Stoyanov, V., & Zettlemoyer, L. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.703 · Cited by 8,821 · cited 1×
“Recently, the development of NLG has conquered the imagination of many thanks to the advances in deep learning technologies, especially the Transformer [159]-based models like BERT [24], BART [83], GPT-2 [127], and GPT3 [13].”
BackgroundLewis, M., Liu, Y., Goyal, N., Ghazvininejad, M., Mohamed, A., Levy, O., Stoyanov, V., & Zettlemoyer, L. (2019) · arXiv · doi:10.48550/arxiv.1910.13461 · Cited by 8,816 · cited 1×
“Recently, the development of NLG has conquered the imagination of many thanks to the advances in deep learning technologies, especially the Transformer [159]-based models like BERT [24], BART [83], GPT-2 [127], and GPT3 [13].”
Background
Research works printed but not linked: 16
- [4] Joshua Albrecht and Rebecca Hwa. 2007. A Re-examination of Machine Learning Approaches for Sentence-Level MT Evaluation. In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics . Association for Computational Linguistics, Prague, Czech Republic, 880-887. https://aclanthology.org/P07-1111 -
- [16] Andrzej Białecki, Robert Muir, Grant Ingersoll, and Lucid Imagination. 2012. Apache lucene 4. In SIGIR 2012 workshop on open source information retrieval . 17. -
- [53] Yilun Du, Shuang Li, Antonio Torralba, Joshua B Tenenbaum, and Igor Mordatch. 2023. Improving Factuality and Reasoning in Language Models through Multiagent Debate. In Forty-first International Conference on Machine Learning . -
- [121] Philipp Koehn and Rebecca Knowles. 2017. Six Challenges for Neural Machine Translation. In First Workshop on Neural Machine Translation . Association for Computational Linguistics, 28-39. -
- [145] Junnan Li, Dongxu Li, Caiming Xiong, and Steven Hoi. 2022. BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation. In ICML . -
- [149] Tian Li, Ahmad Beirami, Maziar Sanjabi, and Virginia Smith. 2020. Tilted Empirical Risk Minimization. In International Conference on Learning Representations . -
- [164] Jiasen Lu, Jianwei Yang, Dhruv Batra, and Devi Parikh. 2018. Neural Baby Talk. In CVPR . -
- [165] Yukun Ma, Khanh Linh Nguyen, Frank Z Xing, and Erik Cambria. 2020. A Survey on Empathetic Dialogue Systems. Information Fusion 64 (2020), 50-70. -
- [175] Marianna J. Martindale, Marine Carpuat, Kevin Duh, and Paul McNamee. 2019. Identifying Fluently Inadequate Output in Neural and Statistical Machine Translation. In MTSummit . -
- [179] Kathleen McKeown. 1992. Text Generation . Cambridge University Press. -
- [209] Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. 2019. Language Models are Unsupervised Multitask Learners. OpenAI Blog 1, 8 (2019), 9. -
- [238] Hui Su, Xiaoyu Shen, Sanqiang Zhao, Zhou Xiao, Pengwei Hu, Cheng Niu, and Jie Zhou. 2020. Diversifying Dialogue Generation with Non-Conversational Text. In 58th Annual Meeting of the Association for Computational Linguistics . ACL, 7087-7097. -
- [256] Zhaopeng Tu, Yang Liu, Lifeng Shang, Xiaohua Liu, and Hang Li. 2017. Neural Machine Translation with Reconstruction. In Thirty-First AAAI Conference on Artificial Intelligence . -
- [261] Neeraj Varshney, Wenlin Yao, Hongming Zhang, Jianshu Chen, and Dong Yu. 2023. A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of LLMs by Validating Low-Confidence Generation. CoRR abs/2307.03987 (2023). https://doi.org/10.48550/arXiv.2307.03987 arXiv:2307.03987 - · arXiv · doi:10.48550/arxiv.2307.03987 · Paper page
- [285] Zeqiu Wu, Michel Galley, Chris Brockett, Yizhe Zhang, Xiang Gao, Chris Quirk, Rik Koncel-Kedziorski, Jianfeng Gao, Hannaneh Hajishirzi, Mari Ostendorf, et al. 2021. A Controllable Model of Grounded Response Generation. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 35. 14085-14093. -
- [291] Xinnuo Xu, Ondřej Dušek, Verena Rieser, and Ioannis Konstas. 2021. AGGGEN: Ordering and Aggregating while Generating. roceedings of the 59th Annual Meeting of the Association for Computational Linguistics (ACL2021) (2021). -
Cited by
Items 1,501–1,600 of 4,844
Foundation Models for Translational Cancer Biology
Tsang, K. K., Kivelson, S., Cortina, J. M. A., Kuchi, A., Berkowitz, J. S., Liu, H., Srinivasan, A., Friedrich, N. A., Fatapour, Y., & Tatonetti, N. P. (2025) · Annual Review of Biomedical Data Science · doi:10.1146/annurev-biodatasci-103123-095633 · Cited by 6
Weber, V., Truelove, A., Gohari, S. S., Wynne, L. A., Kearns, R. J., & O’Carroll, J. (2025) · International Journal of Obstetric Anesthesia · doi:10.1016/j.ijoa.2025.104344 · Cited by 5
Assisting Drafting of Chinese Legal Documents Using Fine-Tuned Pre-trained Large Language Models
Lin, C.-H., & Cheng, P. (2025) · The Review of Socionetwork Strategies · doi:10.1007/s12626-025-00179-5 · Cited by 6
Compound AI System for Personalized Learning: Integrating LLM Agents with Knowledge Graphs
Antonov, E. (2024) · International Conference on Robotics, Intelligent Control and Artificial Intelligence (RICAI) · doi:10.1109/ricai64321.2024.10911764 · Cited by 7
“However, they face challenges [10] such as maintaining long-term context [11], providing reliable source attribution [12], and ensuring interpretability, all of which can undermine trust within educational environments.”
LLM4Jobs: Unsupervised occupation extraction and standardization leveraging Large Language Models
Li, N., Kang, B., & Bie, T. D. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.113302 · Cited by 36
Unmasking Media Bias, Economic Resilience, and the Hidden Patterns of Global Catastrophes
Sufi, F. K., & Alsulami, M. (2025) · Sustainability · doi:10.3390/su17093951 · Cited by 9
Zajac, M., Kulawiak, C., Li, S., Erickson, C., Hubbell, N., & Gong, J. (2025) · Hydrology · doi:10.3390/hydrology12080204 · Cited by 15
“However, a well-documented limitation of LLMs is their tendency to hallucinate, that is, to produce outputs that are syntactically convincing but factually inaccurate or entirely fabricated [90].”
Watson, E. ‘., & Hessami, A. G. (2025) · Electronics · doi:10.3390/electronics14163162 · Cited by 10
Kwon, S., Garodia, R., Lee, M., Yang, Z., & Yu, H. Z. (2023) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2023.acl-long.88 · Cited by 7
Beyer, D., & Cavalcanti, A. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-57259-3 · Cited by 7
Cheok, A. D., & Zhang, E. Y. (2023) · Qeios · doi:10.32388/3ntolq.2 · Cited by 8
Minds: Big questions for linguistics in the age of AI
Backus, A., Cohen, M., Cohn, N., Faber, M., Krahmer, E., Laparle, S., Maier, E., van Miltenburg, E., Roelofsen, F., Sciubba, E., et al. (2023) · Linguistics in the Netherlands · doi:10.1075/avt.00094.bac · Cited by 9
Fine-tuning ChatGPT Achieves State-of-the-Art Performance for Chemical Text MiningPreprint
Zhang, W., Wang, Q., Kong, X., et al. (2023) · ChemRxiv · doi:10.26434/chemrxiv-2023-k7ct5 · Cited by 5
Combining AI and Domain Expertise to Assess Corporate Climate Transition DisclosuresPreprint
Senni, C. C., Schimanski, T., Bingler, J., et al. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4826207 · Cited by 5
“However, LLMs also face two major challenges that arise from relying on internal knowledge: (1) LLMs may produce hallucinated output, that is, text that is not factual 23 , and (2) the embedded knowledge base of LLMs is intransparent and truncated at a certain point in time.”
BackgroundJędrzejczak, W. W., Skarżyńśki, H., & Kochanek, K. (2024) · medRxiv · doi:10.1101/2024.10.08.24315089 · Cited by 6
What are AI hallucinations? Why AIs sometimes make things up
Choi, A., & Mei, K. (2025) · The Conversation · doi:10.64628/aai.etwucf49j · Cited by 8
De la Torre-Trillo, J., Yáñez Castillo, Y., Melgarejo Segura, M. T., Carmona Sánchez, E., Zambudio Munuera, A., Mora-Delgado, J., & López Luque, A. (2026) · Journal of Clinical Medicine · doi:10.3390/jcm15062130 · Cited by 5
“These concerns underscore the necessity of systematic, case-based validation studies before considering integration into real-world clinical workflows [5,6].”
Vakharia, P., Kufeldt, A., Meyers, M., Lane, I., & Gilpin, L. H. (2024) · Lecture Notes in Computer Science · doi:10.1007/978-3-031-71170-1_23 · Cited by 5
“Further, inaccurate model outputs, known as model hallucinations [9], contribute to LLMs’ lack of reliability.”
BackgroundEpisodic memory in AI agents poses risks that should be studied and mitigated
DeChant, C. (2025) · arXiv · doi:10.48550/arxiv.2501.11739 · Cited by 5
Fu, Y., Ramachandran, G. K., Dobbins, N. J., Park, N., Leu, M., Rosenberg, A. R., Lybarger, K., Xia, F., Uzuner, O., & Yetisgen, M. (2024) · arXiv · doi:10.48550/arxiv.2404.00826 · Cited by 5
Zhang, C., Zhu, S., Yang, X., Tseng, Y.-C., Jiang, S., & Rzeszotarski, J. M. (2025) · Proceedings of the 7th ACM Conference on Conversational User Interfaces · doi:10.1145/3719160.3736618 · Cited by 4
Improving Generative Ai Reliability: A Qualitative Study of Hallucination in Large Language Models
Gwon, H. J., Yoo, Y., & Lee, J. Y. (2025) · doi:10.2139/ssrn.5118420 · Cited by 4
Episodic Memory in AI Agents Poses Risks that Should be Studied and Mitigated
DeChant, C. (2025) · 2025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) · doi:10.1109/satml64287.2025.00024 · Cited by 5
“Many contemporary large language models have a great deal of knowledge of the world in general [102]; they can somewhat reliably (hallucinations [103] notwithstanding) answer factual questions about information that was in their training data.”
Li, Y., Fu, X., Verma, G., Buitelaar, P., & Liu, M. (2025) · arXiv · doi:10.48550/arxiv.2510.24476 · Cited by 4
Huang, Q. Y., Luo, Z., Xing, Z., Zeng, J., Chen, J., Xu, X., & Chen, Y. (2024) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3672458 · Cited by 8
Inferring Drug–Gene Relationships in Cancer Using Literature-Augmented Large Language Models
Lai, Y.-J., Wang, L.-J., Yasaka, T. M., Shin, Y., Ning, M., Tan, Y., Shih, C.-H., Guo, Y., Chen, P.-Y., Galloway, H., et al. (2025) · Cancer Research Communications · doi:10.1158/2767-9764.crc-25-0030 · Cited by 4
Traceable Texts and Their Effects: A Study of Summary-Source Links in AI-Generated Summaries
Kambhamettu, H., Flores, J., & Head, A. (2025) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3706599.3719830 · Cited by 16
Comparison of Large Language Model with Aphasia
Watanabe, T., Inoue, K., Kuniyoshi, Y., Nakajima, K., & Aihara, K. (2025) · Advanced Science · doi:10.1002/advs.202414016 · Cited by 6
Mejia, D., Holmes, E. D. V., Marroquin, J., & Benario, J. G. (2025) · ACM Conference on Innovation and Technology in Computer Science Education (ITiCSE) · doi:10.1145/3724363.3729036 · Cited by 6
The Role of AI in Transforming Education: A Systematic Review of Trends
Ratul, Q. E. A., Zaman, E., Das, S. K., & Memari, M. (2025) · Intermountain Engineering, Technology and Computing (IETC) · doi:10.1109/ietc64455.2025.11039329 · Cited by 14
“Furthermore, the “black-box” nature of many AI models complicates transparency, making it difficult for educators and students to trust or critically evaluate AI-generated content [17].”
Zhou, J., Shen, R., You, Y., DiSalvo, C. F., Dombrowski, L. S., & MacLellan, C. J. (2025) · Designing Interactive Systems Conference · doi:10.1145/3715336.3735831 · Cited by 12
Is Having Rationales Enough? Rethinking Knowledge Enhancement for Multimodal Hateful Meme Detection
Lu, J., Xu, B., Zhang, X., Zhu, H., Wang, K., Yang, L., & Lin, H. (2025) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/3726302.3730014 · Cited by 13
Belief in neuromyths among primary school teachers: A cross-national study of 11 countries
Adıgüzel, O. C., Potvin, P., Sarrasin, J. B., Vanhoolandt, C., Corfdir, A., Japashov, N., Mansurova, A., Tsai, C., Wu, C., Elmas, R., et al. (2025) · Trends in Neuroscience and Education · doi:10.1016/j.tine.2025.100264 · Cited by 7
HumanMoD: A multi-RAG collaborative LLM for inclusive urban public healthcare services
Sun, S. M., Zhong, Z., Yu, N., Gong, X., & Yang, K. (2025) · Applied Soft Computing · doi:10.1016/j.asoc.2025.113684 · Cited by 8
Sanchez, C. G., Kharko, A., Hägglund, M., Riggare, S. S., & Blease, C. (2025) · JMIR Research Protocols · doi:10.2196/73602 · Cited by 4
“These tools may also generate dangerous medical advice and can produce incorrect information called “hallucinations” [20,21].”
Rusch, F. R., Ihlow, V. M. J., Schmidt, T., & Mantwill, F. (2025) · Procedia CIRP · doi:10.1016/j.procir.2024.12.107 · Cited by 4
WADSeg: Exploiting weak attention associations for enhanced knowledge segmentation in RAG
Guo, T., Wang, C., Yang, Q., Tang, J., Liu, Y., & Wen, Y. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.129297 · Cited by 7
CLEAR: A Parser-Independent Disambiguation Framework for NL2SQL
Zhang, M., Ma, K., Xu, L., Zhang, K., Peng, Y., & Jin, R. (2025) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde65448.2025.00247 · Cited by 14
Du, J., Wu, J., Kong, D., Liang, W., Hao, F., Xu, J., Wang, B., Wang, G., & Li, P. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3599101 · Cited by 12
“To remedy the hallucination [27], [28], [29] (LVLMs produce incorrect statements to the input images and prompts), compared to getting “hard” prompts (“hard” means prompts from LVLMs are directly input to SAM without any interaction with images), several methods concentrate on deriving “soft”…”
Robillos, R. J. (2025) · Social Sciences & Humanities Open · doi:10.1016/j.ssaho.2025.101886 · Cited by 7
CFA-Bench: Cybersecurity Forensic Llm Agent Benchmark and Testing
Santis, F. D., Huang, K., Valentim, R., Giordano, D., Mellia, M., Houidi, Z. B., & Rossi, D. (2025) · IEEE European Symposium on Security and Privacy Workshops (EuroS&PW) · doi:10.1109/eurospw67616.2025.00031 · Cited by 10
Kalaitzidis, T. J. (2025) · AI & Society · doi:10.1007/s00146-025-02629-8 · Cited by 12
“While much has been written about GenAI’s deep technical limitations (Bender and Gebru 2021; Felin and Holweg 2024; Ji et al. 2023), the significance of these deficits is epistemological: they operate as structural forces that reshape the patterns of cognition when humans engage with GenAI tools.”
Almasri, W., Saad, M., & Mohiyeddini, C. (2025) · Algorithms · doi:10.3390/a18100612 · Cited by 5
From Data‐Driven to Intelligence‐Driven: Innovating Research Paradigms of Perovskite Solar Cells
Li, A., & Yi, C. (2025) · Solar RRL · doi:10.1002/solr.202500681 · Cited by 7
“Additionally, datasets expanded through DFT calculations or generative models exhibit considerable inconsistency [165, 166].”
Zhang, Y. (2025) · Frontiers in Psychology · doi:10.3389/fpsyg.2025.1549755 · Cited by 6
Large Language Models for Service-Oriented Computing (LLM4SOC): Review and Research Directions
Kumara, I., Kaplan, H., Owotogbe, J., Tamburri, D. A., & van den Heuvel, W. (2025) · Communications in computer and information science · doi:10.1007/978-3-032-07313-6_6 · Cited by 6
An Architecture for Integrating Large Language Models with Digital Twins and Automation Systems
Xia, Y., Jazdi, N., & Weyrich, M. (2025) · IEEE International Conference on Emerging Technologies and Factory Automation (ETFA) · doi:10.1109/etfa65518.2025.11205636 · Cited by 10
“Some studies categorize hallucinations as either model-intrinsic or model-extrinsic[13]; however, this distinction alone does not provide actionable guidance for mitigation and solution in terms of system engineering.”
Liu, M., Okuhara, T., Shirabe, R., Nishiie, Y., Xu, Y., Okada, H., & Kiuchi, T. (2025) · JMIR AI · doi:10.2196/78436 · Cited by 6
“The citation itself is real; however, the information it contains does not support the model’s opinion [44].”
Self-citationZhang, Y., Zhang, X., Qi, X., Wu, X., Chen, F., Yang, G., & Fu, H. (2025) · IEEE Reviews in Biomedical Engineering · doi:10.1109/rbme.2025.3619086 · Cited by 17
“However, these models remain critically vulnerable to factual hallucination unless rigorously constrained [215].”
Enhancing Clinical Decision Support with LLMs: A Feasibility Study Integrating CoT, RAG, and QLoRA
Matthew, G., & Hicks, Y. (2025) · Procedia Computer Science · doi:10.1016/j.procs.2025.09.621 · Cited by 5
“Specifically, current LLMs can generate plausible but incorrect information (hallucination [1]), which can pose serious risks to patient safety [2].”
Park, J.-A., Kim, C. H., & Kim, H.-J. (2025) · Telecommunications Policy · doi:10.1016/j.telpol.2025.103106 · Cited by 9
ChatGPT and library users: AI risks of hallucinations and misinformation
Oladokun, B. D., Emmanuel, V. O., Osagie, O., & Alabi, A. O. (2025) · Cybrarians Journal · doi:10.70000/cj.2025.76.642 · Cited by 6
“One prominent example of an AI hallucination involves generating false historical facts (Ji et al., 2023).”
AutoHall: Automated Factuality Hallucination Dataset Generation for Large Language Models
Cao, Z., Yang, Y., Li, X., & Zhao, H. (2025) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/taslpro.2025.3635038 · Cited by 33
“Generally,hallucinationsoflargelanguagemodelscanbe dividedintotwoprimarycategories:intrinsichallucinationsand extrinsichallucinations[5].”
Karam, J., SHUFELT, C. L., Safwan, N., Kapoor, E., Christmas, M., & Faubion, S. S. (2025) · Menopause The Journal of The North American Menopause Society · doi:10.1097/gme.0000000000002695 · Cited by 7
The regulation of fine-tuning: Federated compliance for modified general-purpose AI models
Hacker, P., & Holweg, M. (2025) · Computer law & security review · doi:10.1016/j.clsr.2025.106234 · Cited by 17
Zhang, J., Xia, Y., Zhong, H., Lu, W., Deng, Q., & Wan, C. (2025) · Mathematics · doi:10.3390/math14010090 · Cited by 5
Quan, X. (2026) · Education and Information Technologies · doi:10.1007/s10639-025-13887-w · Cited by 10
Zhang, Y., Qiu, Q., Liu, X., Yu, X., Fu, D., Liu, X., wang, Z., Lin, H., Chen, Y., Yi, L., et al. (2026) · Journal of Optical Communications and Networking · doi:10.1364/jocn.576017 · Cited by 12
Orlando, G. M., Gatta, V. L., Russo, D., & Moscato, V. (2026) · Information Processing & Management · doi:10.1016/j.ipm.2026.104636 · Cited by 5
PAT-Agent: Autoformalization for Model Checking
Zuo, X., Zhang, Y., Wang, H., Cai, Y., Hóu, Z., Sun, J., & Dong, J. S. (2025) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase63991.2025.00176 · Cited by 18
“LLMs are prone to hallucinations and logical inconsistencies [17], [39], especially when generating complex artifacts such as process algebra models or temporal logic specifications.”
Cunningham, E., Kings, I., Robbins, P. T., & Keith, M. J. (2026) · Education for Chemical Engineers · doi:10.1016/j.ece.2026.100503 · Cited by 15
Abdullahi, S., Danyaro, K. U., & Chiroma, H. (2026) · Cluster Computing · doi:10.1007/s10586-025-05891-z · Cited by 16
“Dialogue generation is another key task for LLMs; however the mod - els often produce inaccurate responses, reducing their overall effectiveness [ 16].”
Matharaarachchi, A., Moraliyage, H., Mills, N., Gamage, G., Silva, D. P. D., & Manic, M. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.115469 · Cited by 10
Zhang, Z., Zhao, C., Li, D., Li, P., Chen, Y., Zhu, H., & Han, D. (2026) · Buildings · doi:10.3390/buildings16040719 · Cited by 7
Chen, L., He, R., Lu, P., Jin, Y., Zhou, L., Li, N., Wu, P., & Hu, B. (2026) · Journal of Medical Systems · doi:10.1007/s10916-026-02353-w · Cited by 7
Quantum autopoiesis: A quantum-informational model of recursive identity and knowledge discovery
Stankovski, V. (2026) · Biosystems · doi:10.1016/j.biosystems.2026.105735 · Cited by 4
Use of large language models in undergraduate medical education a scoping review
Uzoechina, G., & Osajiuba, T. (2026) · Discover Education · doi:10.1007/s44217-026-01239-w · Cited by 5
Xu, M., Li, X., Lan, T., Zeng, X., Wang, T., Sheng, S., Deng, X., Sun, J., Fiebrink, R. A., Bryan–Kinns, N., et al. (2026) · Interactive Learning Environments · doi:10.1080/10494820.2025.2604649 · Cited by 6
Enhancing learner-centered feedback with AI: teachers’ practices and perceptions
Aldino, A. A., Maheshi, B., Li, Y., Zhou, Y., Tsai, Y., Gašević, D., & Chen, G. (2026) · Assessment & Evaluation in Higher Education · doi:10.1080/02602938.2026.2638920 · Cited by 9
“…with AI output often involved reducing stylistic excess and modulating tone to preserve authenticity and credibility in their teaching voice (Albadarin et al. 2024; Ji et al. 2023), with broader evidence that the relational/affective dimensions of feedback require human oversight (Dai et al. 2023).”
Lu, L., Luo, P., Sheil, B. B., & Brilakis, I. (2026) · Automation in Construction · doi:10.1016/j.autcon.2026.106896 · Cited by 6
The potential of low-field MRI for global dementia care
Evans, T. E., Harper, J. R., Salehi, A., WEBB, A. G., Cole, J. H., Correa, A., Adams, H. H. H., Cash, D. M., & Jones, D. K. (2026) · Nature Reviews Neurology · doi:10.1038/s41582-026-01199-7 · Cited by 5
“Furthermore, machines such as Swoop embed image enhancement, based on proprietary deep learning architectures, within image reconstruction with little to no access to raw imaging data 76,77 .”
Sevryugina, Y. V., & Vargas, D. (2026) · Journal of Chemical Education · doi:10.1021/acs.jchemed.5c01620 · Cited by 7
Navigating the Risks: A Survey of Security and Privacy Threats in LLM-Based Agents
Gan, Y., Yang, Y., W, Z., He, P., Zeng, R., Wang, Y., Li, Q., Zhou, C., Li, S., Wang, T., et al. (2026) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3807666 · Cited by 122
PRISM: Pipeline for Root-cause Investigation via Specialized Multi-agents
Yan, W., Wu, Y., Liang, P., Yuan, M., Liu, J., Yang, J., & Li, X. (2026) · International Conference on Generative Artificial Intelligence and Information Security (GAIIS) · doi:10.1109/gaiis69281.2026.11519347 · Cited by 15
“Second, models may generate plausible but incorrect diagnoses due to a lack of rigorous fact-checking mechanisms (Hallucination) [3].”
Medical Reasoning With Large Language Models: A Systematic Review and Evaluation
Ren, X., Fan, C., Ma, W., He, H., Gao, C., Zhao, X., & Feng, F. (2026) · iNew Medicine · doi:10.1002/inm3.70056 · Cited by 9
Fan, S., Wang, X., Zhang, B., & Sun, X. (2026) · Telematics and Informatics · doi:10.1016/j.tele.2026.102434 · Cited by 5
TransMap: Pinpointing Mistakes in Neural Code Translation
Wang, B., Li, R., Li, M., & Saxena, P. (2023) · ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) · doi:10.1145/3611643.3616322 · Cited by 16
Fog, Friction, and Failure in Organized Conflict: A Formal Study
Wallace, R. (2024) · Axioms · doi:10.3390/axioms13020111 · Cited by 4
Phantoms of Citation: AI and the Death of the Author-Function
Slater, A. (2024) · Poetics Today · doi:10.1215/03335372-11092818 · Cited by 11
Chui, C. K., Yang, L., & Kao, B. (2024) · ASEE Annual Conference & Exposition · doi:10.18260/1-2--47250 · Cited by 10
Mashagbeh, M. A., Dardas, L. A., Alzaben, H., & Al‐Khayat, A. (2024) · Frontiers in Education · doi:10.3389/feduc.2024.1429324 · Cited by 9
Rose, L., Bewley, S., Payne, M., et al. (2024) · Royal Society Open Science · doi:10.1098/rsos.240698 · Cited by 4
The Development of CanPrompt Strategy in Large Language Models for Cancer Care
Ahmad, N., Mamatjan, E., Wali, T., & Mamatjan, Y. (2024) · IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB) · doi:10.1109/cibcb58642.2024.10702147 · Cited by 4
“Furthermore, the biggest concern in integrating LLMs into healthcare is that they generate plausible-sounding but false results, commonly known as hallucinations [4].”
BackgroundIncitti, F., Salfinger, A., Snidaro, L., & Challapalli, S. S. A. (2024) · International Conference on Information Fusion (FUSION) · doi:10.23919/fusion59988.2024.10706469 · Cited by 9
“Moreover, LLMs’ well-known deficiencies like (iii) their occasional generation of factually incorrect information, known as hallucinations [5], and (iv) the observed variance of produced outputs with respect to slight variations of the input prompt [4], [6], render it an open question whether and…”
BackgroundInfluentialOpanasenko, Y., Bardone, E., Pedaste, M., & Siiman, L. A. (2024) · Education Sciences · doi:10.3390/educsci15010028 · Cited by 6
Examining epistemological challenges of large language models in law
Paseri, L., & Durante, M. (2025) · Cambridge Forum on AI Law and Governance · doi:10.1017/cfl.2024.7 · Cited by 18
“Considering the statistical dimension of LLM’s outcomes, pieces of false knowledge (the so-called hallucination, Ji et al., 2023; Rebuffe et al., 2022) are frequently generated however plausible: “Readers and consumers 8 See, for instance, Frank Fagan, who envisages that “legal tasks will take less…”
GigSense: An LLM-Infused Tool for Workers’ Collective Intelligence
Imteyaz, K., Flores-Saviaga, C., & Savage, S. (2024) · Avances en Interacción Humano-Computadora · doi:10.47756/aihc.y9i1.159 · Cited by 11
stEELlm: An LLM for Generating Semantic Annotations of Tabular Data
Cremaschi, M., D’Adda, F., & Maurino, A. (2026) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3719206 · Cited by 11
“Further studies reveal that LLMs are not able to recall facts and often experience hallucinations by generating statements that are factually incorrect [42, 104].”
Bogers, T., Gäde, M., Hall, M. M., Koolen, M., Petras, V., & Skov, M. (2025) · Conference on Human Information Interaction and Retrieval · doi:10.1145/3698204.3716466 · Cited by 6
Ahmad, P. N., Ullah, I., Aboelenein, N. M., Singh, S. K., Jiang, W., Al‐Khasawneh, M. A., & Daradkeh, Y. I. (2025) · ACM Transactions on Internet of Things · doi:10.1145/3744754 · Cited by 10
Personalized News Recommendation Towards the Era of LLMs: Review and Prospect
Li, J., Liu, Z., Hu, L., Rao, Y., Liu, B., Fang, B., & Nie, L. (2025) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2025.3581806 · Cited by 6
Pei, F., Lin, Y., Liu, J., Zhuang, C., & Zhai, S. (2025) · Chinese Journal of Mechanical Engineering · doi:10.1186/s10033-025-01275-x · Cited by 7
“However, LLMs still struggle to distinguish between true and false information in the input, and may generate misleading guidance due to hallucinations [15], thereby requiring extensive human expert review of their outputs.”
Model Editing for LLMs4Code: How Far are we?
Li, X., Wang, S., Li, S., Ma, J., Yu, J., Liu, X., Wang, J., Ji, B., & Zhang, W. (2025) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse55347.2025.00049 · Cited by 23
“This inevitably introduces a significant amount of noisy information into the training data, which can potentially result in incorrect outputs from LLMs4Code models [16].”
Schmid, C., Kastner, F., Zhang, D., et al. (2025) · Journal of Industrial Ecology · doi:10.1111/jiec.70058 · Cited by 4
Russinovich, M., Salem, A., Zanella-Béguelin, S., & Zunger, Y. (2025) · Communications of the ACM · doi:10.1145/3749447 · Cited by 10
Applications of GPT in Political Science Research: Extracting Information from Unstructured Text
Lee, K., Paci, S., Park, J., You, H. Y., & Zheng, S. (2025) · PS Political Science & Politics · doi:10.1017/s1049096525000046 · Cited by 8
“…GPT ’ s strengths in natural-language processing while mitigating its weaknesses in complex reasoning and “ hallucination ” (i.e., false information) (Ji et al. 2023; Wei et al. 2022) along with the reliability and consistency of synthetic survey data produced by LLMs (Bisbee et al., 2024).”
González-Cantera, T., & Bonacchi, C. (2025) · International Journal of Heritage Studies · doi:10.1080/13527258.2025.2543736 · Cited by 4
“This is because GenAI can experience a phenomenon called ‘hallucination’ (Alkaissi and McFarlane 2023; Ji et al. 2023), generating incorrect or misleading answers.”
Refining medical large language models: key insights from instruction tuning
Alqahtani, M. Q., Al-Barakati, A. A., & Alotaibi, F. S. (2025) · PeerJ Computer Science · doi:10.7717/peerj-cs.3216 · Cited by 7
Oyserman, D., Lai, M. H. C., Burbidge, A., & Sorensen, N. (2025) · The Journal of Experimental Education · doi:10.1080/00220973.2025.2565170 · Cited by 4
Das, A. B., Ahmed, S., & Sakib, S. K. (2025) · IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI ...) · doi:10.1109/bhi67747.2025.11269492 · Cited by 16
“Hallucinations: LLMs are known to generate plausible yet incorrect or non-existent information, a phenomenon referred to as hallucination [10], [11], [12].”
Metadata
- Authors
- Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., Ishii, E., Bang, Y. J., Madotto, A., & Fung, P.
- Affiliations
- Hong Kong University of Science and Technology (HK)
- ORCID
- Venue
- ACM Computing Surveys
- Published
- 2023-03-03
- Volume
- 55
- Issue
- 12
- Pages
- 1-38
- ISSN
- 0360-0300 (ISSN-L)
- 1557-7341
- OpenAlex
- W4309674289
- Semantic Scholar
- 3def68bd0f856886d34272840a7f81588f2bc082
- Cited by
- 5,025
- References
- 111
- Citations by year
- 2026: 2296 · 2025: 1441 · 2024: 817 · 2023: 453
- FWCI
- 224.36
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- ACM Computing Surveys (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Rare & Special e-Zone (The Hong Kong University of Science and Technology) (submittedVersion)
- arXiv (Cornell University) (cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Speech and dialogue systems
- Keywords
- hallucinations, natural language generation, hallucination mitigation, abstractive summarization, dialogue generation, generative question answering, data-to-text generation, machine translation, large language models, transformer models
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-09-30, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-09-30 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.




