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 1,701–1,800 of 4,844
A Review of Explainable AI for Android Malware Detection and Analysis
Tanha, M., & Kafaie, S. (2025) · IEEE Access · doi:10.1109/access.2025.3597575 · Cited by 11
“However, standalone LLMs may hallucinate incorrect or unverifiable mappings if not grounded in authoritative knowledge sources [63].”
Sun, T., Chan, G., Stjepanović, D., Mekonen, T., Vu, G. T., Lim, C. C. W., McClure‐Thomas, C., Russel, C., Connor, J. P., Hall, W., et al. (2025) · Drug and Alcohol Review · doi:10.1111/dar.70022 · Cited by 4
“These include factual inaccuracies or hallucinations [17], outdated information due to AI models' lack of real-time updates [14, 18], and biased content that may exclude or misrepresent diverse groups, as current AI systems are primarily trained on vast internet-sourced datasets that reflect…”
Wang, J.-F., Chan, D.-Y., Tsai, H., & Fang, B.-X. (2025) · Data & Knowledge Engineering · doi:10.1016/j.datak.2025.102507 · Cited by 4
Onwuegbuche, F. C., Titung, R., Rantanen, E. M., Jurcut, A. D., Alm, C. O., & Pasquale, L. (2025) · IEEE Access · doi:10.1109/access.2025.3615788 · Cited by 10
“LLMs occasionally generate persuasive yet inaccurate or contextually irrelevant outputs, referred to as hallucinations [98].”
Fine-tuning large language models in federated learning with fairness-aware prompt selection
Jiang, Y., Li, Z., & Song, B. (2025) · Neural Networks · doi:10.1016/j.neunet.2025.108160 · Cited by 5
Asking Questions to Alleviate Object Hallucination in Large Vision-Language Models
Bi, C., Dang, T., Wang, S., Cao, F., & Huang, Q. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3618949 · Cited by 3
“Therefore, many studies have been carried out with the aim of reducing the impact of hallucinations [29], [30], [31].”
Demystifying Feature Requests: Leveraging LLMs to Refine Feature Requests in Open-Source Software
Pragyan, K. C., Ghandiparsi, R., Herron, T., Heaps, J., & Hosseini, M. B. (2025) · IEEE International Requirements Engineering Conference (RE) · doi:10.1109/re63999.2025.00020 · Cited by 5
“However, studies show these models often hallucinate when processing text with NL defects [72].”
Sorour, S. E., & Almusallam, N. (2025) · Alexandria Engineering Journal · doi:10.1016/j.aej.2025.09.070 · Cited by 5
Hallucination in Large Language Models: From Mechanistic Understanding to Novel Control Frameworks
Zhang, W., Zhang, C., Gu, C., Kou, J., Yuan, H., Fang, X., Duan, X., & Fang, Y. (2024) · International Conference on Universal Village (UV) · doi:10.1109/uv63228.2024.11189226 · Cited by 60
“Multilingual transfer effects : In multilingual models, knowledge encoded in one language may transfer imperfectly to others, creating language specific hallucinations that reflect translation ambiguities or cultural misalignments [8].”
InfluentialAlyukov, M., Makhortykh, M., Voronovici, A., & Sydorova, M. (2025) · Harvard Kennedy School Misinformation Review · doi:10.37016/mr-2020-187 · Cited by 21
Whitty, M. T. (2025) · Electronics · doi:10.3390/electronics14204104 · Cited by 6
Large Language Models for Structured Information Processing in Construction and Facility Management
Buga, K., Tesic, R., Koyuncu, E., & Hanne, T. (2025) · Electronics · doi:10.3390/electronics14204106 · Cited by 8
Bridging physical constraints and deep generative models via physics-aware normalizing flows
Benjamin, S., Mendikowski, M., & Schmid, T. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.131880 · Cited by 3
Research on the methodology of personalized recommender systems based on multimodal knowledge graphs
Bao, S., & Wang, J. (2025) · Natural Language Processing Journal · doi:10.1016/j.nlp.2025.100193 · Cited by 6
Multi-Source Knowledge Pruning for Retrieval-Augmented Generation: A Benchmark and Empirical Study
Yu, S., Cheng, M., Liu, Q., Wang, D., Yang, J., Ouyang, J., Luo, Y., Lei, C., & Chen, E. (2025) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3746252.3761340 · Cited by 24
“Nevertheless, since LLMs rely solely on internal knowledge acquired during training, they are often susceptible to hallucination (Zhou et al., 2020; Ji et al., 2023a,b; Mallen et al., 2022).”
A Framework for Automation in Psychotherapy
Imel, Z. E., Creed, T. A., Kious, B. M., Althoff, T., Atzil–Slonim, D., & Srikumar, V. (2025) · Current Directions in Psychological Science · doi:10.1177/09637214251386047 · Cited by 7
“At least currently, generative language models may sometimes appear to reflect true knowledge about the world, but statements can be incorrect or made up completely—a phenomenon often referred to as “hallucinations” (Ji et al., 2023).”
Engelstein, H., Ramon‐Gonen, R., Barbash, I. M., Beinart, R., Cohen, M. S., & Sabbag, A. (2025) · Journal of Medical Systems · doi:10.1007/s10916-025-02289-7 · Cited by 2
“This is especially important given the risk of hallucinations and overconfident misclassifications that LLMs may produce when faced with ambiguous or unfamiliar inputs [ 36].”
Survey of GenAI for Automotive Software Development: From Requirements to Executable Code
Petrović, N., Zolfaghari, V., Schamschurko, A., Kirchner, S., Pan, F., Wu, C., Purschke, N., Velsh, A., Lebioda, K., Song, Y., et al. (2025) · International Generative AI and Computational Language Modelling Conference (GACLM) · doi:10.1109/gaclm67198.2025.11232000 · Cited by 8
“Following [38], hallucinations are generated content that is nonsensical or unfaithful to the provided source content.”
CHAIR-Classifier of Hallucination As Improver
Sun, A. (2025) · doi:10.1109/ijcnn64981.2025.11227344 · Cited by 3
“Hallucinations are a common issue in large language models (LLMs), limiting their reliability in critical applications, and a lot of research has been working on it [1] [2].”
Eminovic, S., Lindholz, M., Dell’Orco, A., Meddeb, A., Baumgaertner, G. L., Schulze‐Weddige, S., Siebert, E., Wattjes, M. P., Penzkofer, T., & Nawabi, J. (2025) · European Radiology · doi:10.1007/s00330-025-12122-x · Cited by 4
Sharma, N., & Singh, S. (2025) · Journal of Transformative Technologies and Sustainable Development · doi:10.1007/s41314-025-00083-4 · Cited by 6
Wang, S., Yan, R., & Hu, W. (2025) · Education and Information Technologies · doi:10.1007/s10639-025-13839-4 · Cited by 3
“…the new generation of Gen AI systems creates data illusions and that their immediate benefits often result in an underestimation of both short-term and long-term negative impacts (Saúde et al., 2024), particularly adversely affecting students’ understanding and learning outcomes (Ji et al., 2023).”
Goswami, Y., Bangde, S., Singh, S., & Khosla, A. (2025) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2025.111231 · Cited by 6
Oedipus: LLM-enchanced Reasoning CAPTCHA Solver
Deng, G., Ou, H., Liu, Y., Zhang, J., Zhang, T., & Liu, Y. (2025) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/3719027.3744872 · Cited by 40
Cho, N., Li, H., Zhang, X., & Li, Y. (2025) · Journal of Science Education and Technology · doi:10.1007/s10956-025-10280-7 · Cited by 8
“There are some concerns about the quality and correctness of the responses, such as “hallucination” (Ji et al., 2023).”
Damo, G., Ocampo, N. B., Cabrio, E., & Villata, S. (2025) · Data & Knowledge Engineering · doi:10.1016/j.datak.2025.102535 · Cited by 4
A hallucination detection and mitigation framework for faithful text summarization using LLMs
Liu, S., Gao, Y., Li, S., Wang, P., & Wang, T. (2025) · Scientific Reports · doi:10.1038/s41598-025-31075-1 · Cited by 16
Chain-of-Thought Prompt Optimization via Adversarial Learning
Yang, G., Cai, X., Wang, S., & Liu, J. (2025) · Information · doi:10.3390/info16121092 · Cited by 3
Beyond sparsity: an empirical study of structured collaboration in modular AI
Zhou, X., Kim, S., Wang, Y., & Zhang, K. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.131616 · Cited by 4
Adaptive Retrieval-Augmented Embodied Intelligence RAG Reasoning Framework
Chen, M., Bai, Y., Fang, X., & Yi, P. (2025) · IEEE International Conference on Software Engineering and Artificial Intelligence (SEAI) · doi:10.1109/seai65851.2025.11108876 · Cited by 3
“Studies indicate that in multimodal tasks, misalignment between LLM-generated text and visual inputs leads to significant reasoning errors [3] .”
Evolution of Reference Services in the Era of Generative Artificial Intelligence
Aguiñaga, J., Mooradian, N., Ghosh, S., & Hofman, D. L. (2025) · Proceedings of the Association for Information Science and Technology · doi:10.1002/pra2.1232 · Cited by 7
Lutz, C., Newlands, G., & Jarrahi, M. H. (2025) · Springer eBooks · doi:10.1007/978-981-97-8440-0_87-1 · Cited by 8
Comprehensive Scoping Review on Discrepancy in Accuracy of ChatGPT in Dental Health Practice
Bommanavar, S., Varma, S. R., & Karobari, M. I. (2025) · Human Behavior and Emerging Technologies · doi:10.1155/hbe2/1269498 · Cited by 4
Large Language Models and Machine Learning Framework for Predicting Dental Ceramics Performance
Zhou, H., Bai, Y., Chen, Y., Fan, D., Wang, P., Ji, P., & Chen, T. (2025) · International Dental Journal · doi:10.1016/j.identj.2025.109358 · Cited by 5
“44 Moreover, the application of LLMs in medical literature mining must be approached with caution; rigorous expert oversight is essential to ensure accuracy and to prevent biased or erroneous outputs.”
Qi, J., Zhao, H., Xiang, R., Zhang, Y., Zhang, X., Wang, C., Jian, L., & Ji, F. (2026) · Journal of Research on Technology in Education · doi:10.1080/15391523.2025.2601067 · Cited by 9
A Survey of Hallucination in Large Language Models
何, 金. (2026) · Artificial Intelligence and Robotics Research · doi:10.12677/airr.2026.151016 · Cited by 5
Vidyarthi, A., Singh, M. K., & Moirangthem, D. S. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.131160 · Cited by 7
ChatGPT as economics tutor: Capabilities and limitations
Bröse, N., Spielmann, C., & Tode, C. (2026) · International Review of Economics Education · doi:10.1016/j.iree.2025.100337 · Cited by 7
“Hallucinations are particularly problematic because they are often difficult to detect; the generated text is typically fluent and well-structured, giving an impression of accuracy (Ji et al. 2023).”
Chen, M., Wang, Y., Ai, L., Cheng, X., Jh, X., Yi, H., Lan, E., Liu, J., Luo, J., Qiu, Y., et al. (2026) · IEEE Transactions on Consumer Electronics · doi:10.1109/tce.2026.3653798 · Cited by 5
“These models often su ff er from significant domain knowledge gaps and a propensity for hallucinations, that is, generating plausible but factually incorrect outputs, rendering them untrustworthy for tasks that demand strict adherence to physical laws and engineering principles [12], [13], [14], [15].”
Aziz, I., Arshad, S., & Hussan, M. A. (2026) · Asian Journal of Psychiatry · doi:10.1016/j.ajp.2026.104851 · Cited by 6
Trustworthiness of large language models: hallucinations
Brunello, N. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-33592-1.00007-3 · Cited by 5
How individuals use generative AI for personal financial management
Pak, T. (2026) · Journal of Behavioral and Experimental Finance · doi:10.1016/j.jbef.2026.101145 · Cited by 3
GastroTCM: a large language model assistant for gastroenterology in traditional Chinese medicine
Wang, L., Tang, K., Yang, Z., Wang, Y., Zhang, P., Wu, B., Zhao, W., Chen, J., Gong, J., Du, S., et al. (2026) · Chinese Medicine · doi:10.1186/s13020-025-01295-8 · Cited by 10
“However, recent evaluations have shown that general-purpose LLMs still make frequent and sometimes clinically important errors when answering medical questions, raising concerns about their safe deployment in real-world care settings [10, 11].”
Fair Foundation Models for Medical Image Analysis: Challenges and Perspectives
Queiroz, D., Carlos, A., Anjos, A., & Berton, L. (2025) · ACM Trans. Comput. Heal. · doi:10.1145/3793542 · Cited by 11
“They are typically associated with high uncertainty and incomplete or inaccurate knowledge [97, 98, 99, 100], particularly in instances where the model lacks relevant training data and exhibits low confidence [101, 102, 97].”
Hu, C., Huang, Y., Kuang, J., Dai, B., Peng, Y., Xiao, Y., & Su, Y. (2026) · Electronics · doi:10.3390/electronics15030541 · Cited by 3
BonsAIDE: An Extended Vision for Human–AI Interaction in IDEs
Moreno-Lumbreras, D., Kula, R. G., & Treude, C. (2026) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3793681 · Cited by 3
LLM-Powered Multi-Agent Collaboration for Intelligent Industrial On-Call Automation
Fu, R., Zhang, Y., Che, Z., Wu, X., Zhong, Z., Ren, Z., Zhang, S., Wang, F., Sun, Y., Liu, X., et al. (2025) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase63991.2025.00184 · Cited by 8
“Inputting raw tickets directly into an LLM may cause the model to hallucinate due to noisy inputs, overly relying on the beginning and end of the text while neglecting important information in the middle [25], [26].”
Predictable artificial intelligence
Zhou, L., Casares, P. A. M., Martínez‐Plumed, F., Burden, J., Burnell, R., Cheke, L. G., Ferri, C. S., Marcoci, A., Mehrbakhsh, B., Moros-Daval, Y., et al. (2026) · Artificial Intelligence · doi:10.1016/j.artint.2026.104491 · Cited by 42
Tang, Z., Su, Y., Hansen, P., & Zhang, P. (2026) · Information Processing & Management · doi:10.1016/j.ipm.2026.104668 · Cited by 5
Quantifying Factual Divergence in Generative Models: SHAP-LIME Based Hallucination Score for LLMs
Haq, I., Saqib, M., Zhang, Y., & Khan, I. A. (2026) · Multimedia Systems · doi:10.1007/s00530-025-02150-4 · Cited by 4
“Ji et al. [6] systematically categorized hallucinations as intrinsic, caused by internal biases and miscalibrated probabilistic reasoning, and extrinsic, resulting from incomplete contextual grounding or missing factual links to external knowledge sources.”
Nalvakku: fact-based counter narratives for Tamil hate speech
Ravikumar, N., Kumaresan, P. K., Saranya, R., Navaneethakrishnan, S. C., & Chakravarthi, B. R. (2026) · International Journal of Data Science and Analytics · doi:10.1007/s41060-026-01023-x · Cited by 3
When AI reviews science: Can we trust the referee?
Wang, J., Liu, Y., Xu, H., Hu, K., Di, S., Ni, W., Yue, L., Zhang, M.-L., Ren, K., & Chen, L. (2026) · The Innovation Informatics · doi:10.59717/j.xinn-inform.2026.100030 · Cited by 10
“Other identified pitfalls of LLM-based referees include factual errors (hallucinations) and a range of cognitive biases that undermine trust in AI judgments 11 .”
What is a fact in the age of generative AI? Fact-checking as an epistemological lens
Dierickx, L., Opdahl, A. L., Bjerknes, F., & Lindén, C. (2026) · Information Communication & Society · doi:10.1080/1369118x.2026.2630697 · Cited by 10
“Second, they are prone to generating semantic noise, ranging from subtle misinterpretations to entirely fabricated but plausible content – a phenomenon commonly referred to as ‘hallucination’ (Ji et al., 2023; Perković et al., 2024).”
Souza, F. A., & Camboim, S. P. (2026) · ISPRS International Journal of Geo-Information · doi:10.3390/ijgi15020087 · Cited by 3
Hallucination in LLM-Based Code Generation: An Automotive Case Study
Pavel, M., Petrović, N., Mazur, L., Zolfaghari, V., Pan, F., & Knoll, A. (2025) · International Conference on Foundation and Large Language Models (FLLM) · doi:10.1109/fllm67465.2025.11391125 · Cited by 7
Efficient Privacy-Preserving Retrieval Augmented Generation with Distance-Preserving Encryption
Ye, H., Guo, J., Liu, Z., & Lam, K. (2025) · International Conference on Foundation and Large Language Models (FLLM) · doi:10.1109/fllm67465.2025.11391120 · Cited by 11
“…Models (LLMs), despite their powerful capabilities in generation and reasoning across various domains, face well-known limitations such as hallucination, outdated knowledge due to static training corpora, and lack of access to proprietary or time-sensitive information during inference [1], [2].”
Marlowe, T. J., Ku, C. S., Laracy, J. R., Kirova, V. D., & Herbert, K. G. (2026) · Journal of Integrated Design and Process Science · doi:10.1177/10920617251405471 · Cited by 13
“Moreover, while AI can process and analyze vast amounts of data, it lacks true understanding and cannot replicate the nuanced judgment of human educators (Ji et al., 2023; Marcus & Davis, 2020).”
Shah, S. T. U., Hussein, M., Barcomb, A., Moshirpour, M., & Leung, H. (2026) · Information and Software Technology · doi:10.1016/j.infsof.2026.108066 · Cited by 5
Lee, H., & Lim, S. (2026) · Applied Sciences · doi:10.3390/app16052244 · Cited by 5
Evaluating the performance of ChatGPT-5, Claude, and DeepSeek for statistical analysis in medicine
Sebo, P., & Wang, T. (2026) · Internal and Emergency Medicine · doi:10.1007/s11739-026-04291-4 · Cited by 6
MedFusionT5: Cross-Modal Attention Boosts Semantic Quality and Reduces Hallucinations in Dental AI
Abdaoui, H., Barbaria, S., Dergaa, I., Ceylan, H. İ., Bragazzi, N. L., de Giorgio, A., Salah, R. B., & Rahmouni, H. B. (2026) · International Dental Journal · doi:10.1016/j.identj.2025.109404 · Cited by 4
“29,30 Although cross-attention mechanisms have proven effective in general vision-language tasks, 31,32 their application in medical report generation remains underexplored.”
Hong, S. J. (2026) · Information Communication & Society · doi:10.1080/1369118x.2026.2634845 · Cited by 3
“Large language models can produce confident yet inaccurate responses; Ji et al. (2023) identify hallucination – plausible but unsupported content generation – as a persistent limitation of neural language generation systems.”
RAG Security and Privacy: Formalizing the Threat Model and Attack Surface
Arzanipour, A., Behnia, R., Ebrahimi, R., & Dutta, K. (2025) · IEEE ... International Conference on Data Mining workshops · doi:10.1109/icdmw69685.2025.00165 · Cited by 26
“By mitigating issues like hallucination [9], factual inconsistency, and limited generalization [10] that often affect standalone LLMs, RAG architectures offer a promising foundation for building more reliable and knowledge-aware AI systems.”
Text Mining of CVD Synthesis Recipes for 2D Materials
Lu, A., Chen, R., Yao, A., Chen, M.-C., Park, J. H., Zhang, T., Zheng, X., Mao, N., Wang, J., Wang, Z., et al. (2026) · Advanced Materials · doi:10.1002/adma.202509132 · Cited by 6
“The stepwise nature of our framework allows us to independently inspect model outputs after each step and mitigate model prediction errors through further finetuning or rule-based corrections, in contrast to the low interpretability and risk of hallucination of LLMs [11].”
Huynh, L., & McNamara, D. S. (2026) · Electronics · doi:10.3390/electronics15061262 · Cited by 7
From complexity to clarity: GPT-4 for patient-centered prostate cancer pathology reports
Fereydooni, S., Homer, R., & Chartash, D. (2026) · Patient Education and Counseling · doi:10.1016/j.pec.2026.109597 · Cited by 4
Predicting Retrieval Utility and Answer Quality in Retrieval-Augmented Generation
Tian, F., Ganguly, D., & Macdonald, C. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-21289-4_24 · Cited by 16
A Survey on Large Language Models in Software Security: Opportunities and Threats
Rashid, M. B., Hossain, M. S. J., Khan, M. I., Tahora, S., Siddika, A., Prakash, M. I., Yeasmin, S., & Shahriar, H. (2026) · Computers · doi:10.3390/computers15040226 · Cited by 9
Mohamed, R. A. A. (2026) · Interactive Learning Environments · doi:10.1080/10494820.2026.2649549 · Cited by 4
“…educational risks: hallucinated or overconfident feedback can foster automation bias and distort epistemic calibration (Bender et al., 2021; Ji et al., 2023; Parasuraman & Riley, 1997), while easy text generation raises integrity and authorship concerns (Draxler et al., 2023; Fleckenstein…”
InfluentialSecurity and Privacy of Large Language Models: Threat Taxonomy, Ethical Implications, and Governance
Pribisalić, M., & Martinčić-Ipšić, S. (2026) · AI · doi:10.3390/ai7050152 · Cited by 9
A Survey on Fine-Grained Multimodal Large Language Models
Peng, Y., Wang, Z., Li, G., Zheng, X., Yin, S., & He, H. (2026) · Chinese Journal of Electronics · doi:10.23919/cje.2025.00.336 · Cited by 34
Binding events, building meaning: A new framework for pragmatic interpretation
Shi, E. R. (2026) · Journal of Neurolinguistics · doi:10.1016/j.jneuroling.2026.101359 · Cited by 6
OpenVid++: A Large-scale High-quality Dataset for Text-to-video Generation
Nan, K., Fan, T., Xie, R., Zhou, P., Zhang, Y., Li, X., Yang, J., Yang, Z., & Tai, Y. Y. (2026) · International Journal of Computer Vision · doi:10.1007/s11263-026-02989-8 · Cited by 8
“…intrinsic hallucinations (content directly contradicting source video information) and extrinsic hallucinations (unverifiable content, including extrinsic facts consistent with general knowledge but absent from the source video, and extrinsic non-facts) (Wang et al., 2024; Ji et al., 2023).”
Du, Y., & Yuan, Y. (2026) · Assessment & Evaluation in Higher Education · doi:10.1080/02602938.2026.2732250 · Cited by 13
Buhl, D., Szafarski, D., Welz, L., & Lanquillon, C. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-35894-4_3 · Cited by 4
Evaluating and Modeling Attribution for Cross-Lingual Question Answering
Müller, B., Wieting, J., Clark, J. H., Kwiatkowski, T., Ruder, S., Soares, L. B., Aharoni, R., Herzig, J., & Wang, X. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.10 · Cited by 26
Walking a Tightrope – Evaluating Large Language Models in High-Risk Domains
Hung, C.-C., Rim, W. B., Frost, L., Bruckner, L., & Lawrence, C. (2023) · GenBench Workshop on Generalisation (Benchmarking) in NLP (GenBench) · doi:10.18653/v1/2023.genbench-1.8 · Cited by 27
Epistemic modality in English and Thai medical abstracts: a corpus-based study
Satthachai, M., & Kenny, D. (2024) · Perspectives · doi:10.1080/0907676x.2023.2300042 · Cited by 3
“…one of the consequences of the world’s current fascination with artificial intelligence (AI) is likely to be a growing concern about the often-misplaced confidence with which AI systems generate what turn out to be the wholly unmoti-vated inaccuracies known as ‘hallucinations’ (Ji et al., 2023).”
Developers’ Perspective on Trustworthiness of Code Generated by ChatGPT: Insights from Interviews
Rabani, Z. S., Khorashadizadeh, H., Abdollahzade, S., Groppe, S., & Ghofrani, J. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-55486-5_16 · Cited by 4
What Should a Robot Do? Comparing Human and Large Language Model Recommendations for Robot Deception
Rogers, K., Webber, R. J. A., Zubizarreta, G. G., Cruz, A. M., Chen, S., Arkin, R. C., Borenstein, J., & Wagner, A. R. (2024) · ACM/IEEE International Conference on Human-Robot Interaction (HRI) · doi:10.1145/3610978.3640752 · Cited by 4
Research on confusing responses based on ChatGPT
Xiong, H. (2024) · Applied and Computational Engineering · doi:10.54254/2755-2721/57/20241315 · Cited by 3
Unraveling Knowledge-Based Chatbot Adoption Intention in Enhancing Species Literacy
Manik, L. P., Rini, D. S., Priyanti, P., Indrawati, A., Fefirenta, A. D., Akbar, Z., Sumowardoyo, T. D., Apriani, N. F., & Kartika, Y. A. (2024) · Interdisciplinary Journal of Information Knowledge and Management · doi:10.28945/5280 · Cited by 7
“Some studies reported a need for more trust in chatbots because sometimes they generate hallucinated responses (Alkaissi & McFarlane, 2023; Ji et al., 2023).”
BackgroundCommentVis: Unveiling Comment Insights Through Interactive Visualization Tool
Yan, G., Jang, J., & Seo, J. (2024) · IEEE Pacific Visualization Conference (PacificVis) · doi:10.1109/pacificvis60374.2024.00050 · Cited by 3
“While GPT-4-turbo is famous for its good performance in text generation, it is also well-known for generating ‘hallucina-tions’: text that is nonsensical or unfaithful to the provided source input [12].”
BackgroundCivilSum: A Dataset for Abstractive Summarization of Indian Court Decisions
Malik, M., Zhao, Z., Fonseca, M., Rao, S., & Cohen, S. B. (2024) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/3626772.3657859 · Cited by 16
James, J. K., Maran, T., Rice, M. P., Hunt, T. S., Peterson, K. J., Hogan, W. J., Damani, S., & Ryu, A. J. (2024) · Mayo Clinic Proceedings Digital Health · doi:10.1016/j.mcpdig.2024.08.001 · Cited by 11
“However, this requires further work to ensure summariza-tions are accurate and free of hallucinations 11 and bias.”
Rejithkumar, G., Anish, P. R., & Ghaisas, S. S. (2024) · IEEE International Requirements Engineering Conference (RE) · doi:10.1109/re59067.2024.00037 · Cited by 6
Self-Directed Learning for Community Health Workers in Malawi Through Generative AI
Taylor, A., Magwira, M., Chamangwana, C., Chapuma, E., Liwewe, T., & Kankhwali, C. (2024) · IEEE International Conference on Healthcare Informatics (ICHI) · doi:10.1109/ichi61247.2024.00092 · Cited by 3
“We are involving experts to develop and review questions and answers and to adapt metrics [28] to measure the quality of the answers and questions.”
BackgroundAsynchronous Event Driven Brain Teaser Using Node.js
Subbulakshmi, N., Begum, S. A., Vignesh, R. V., & Chandru, R. (2024) · International Conference on Electronics and Sustainable Communication Systems (ICESC) · doi:10.1109/icesc60852.2024.10689905 · Cited by 4
“This problem arises when GPTs produce literature that is coherent and purportedly realistic but has factual errors, contradictions, or content that is not in line with human cultural norms and expectations [4], [5].”
BackgroundŁajewska, W., Balog, K., Spina, D., & Trippas, J. R. (2024) · Annual International ACM SIGIR Conference on Research and Development in Information Retrieval in the Asia Pacific Region (SIGIR-AP) · doi:10.1145/3673791.3698409 · Cited by 11
Chernyshova, I. A., Васильева, И. А., Moor, N. A., Ivanisenko, N. V., Kutuzov, M. M., Abramova, T. V., Zakharenko, A. L., & Lavrik, O. I. (2024) · International Journal of Molecular Sciences · doi:10.3390/ijms252312526 · Cited by 3
Toward the Clinically Effective Evaluation of Artificial Intelligence–Generated Responses
Anbumani, S., & Ahunbay, E. E. (2024) · JCO Clinical Cancer Informatics · doi:10.1200/cci-24-00258 · Cited by 4
Wu, J., Zhou, X., Kuligowski, E. D., & Zhang, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-981-96-1242-0_2 · Cited by 3
Large Vision-Language Model Security: A Survey
Wang, T., Fang, Z., Xue, H., Zhang, C., Jin, M., Xu, W., Shu, D., Yang, S. J., Wang, Z. G., & Liu, D. (2024) · Communications in computer and information science · doi:10.1007/978-981-96-0151-6_1 · Cited by 13
Chatbot Dialog Design for Improved Human Performance in Domain Knowledge Discovery
Oruche, R., Cheng, X., Zeng, Z., Vazzana, A., Goni, M. A., Shibo, B. W., Goruganthu, S. K., Kee, K. F., & Calyam, P. P. (2025) · IEEE Transactions on Human-Machine Systems · doi:10.1109/thms.2024.3514742 · Cited by 3
“In addition, while chatbot dialogs are able to provide responses for domain users that guide them in knowledge discovery tasks, generative models such as LLMs have shown to be prone to hallucination [65] by providing falsehoods or misspeaking on user intents.”
Img2Vocab: Explore Words Tied to Your Life With LLMs and Social Media Images
Yamaoka, K., Watanabe, K., Kise, K., Dengel, A. R., & Ishimaru, S. (2025) · IEEE Access · doi:10.1109/access.2025.3533076 · Cited by 4
“With the large parameters and training data, GPT-3 is often capable of generating natural text generation, although it has room for improvements in multiple aspects, including ‘‘hallucination,’’ where a language model generates nonsensical or unfaithful output given an input [17].”
Signalling and social learning in swarms of robots
Cazenille, L., Toquebiau, M., Lobato‐Dauzier, N., Loi, A., Macabre, L., Aubert-Kato, N., Genot, A. J., & Bredèche, N. (2025) · Philosophical Transactions of the Royal Society A Mathematical Physical and Engineering Sciences · doi:10.1098/rsta.2024.0148 · Cited by 14
Straight Outta Logs: Can Large Language Models Overcome Preprocessing in Next Event Prediction?
Brennig, K., Kaltenpoth, S., & Müller, O. (2025) · Lecture notes in business information processing · doi:10.1007/978-3-031-78666-2_15 · Cited by 3
Ye, H., Xu, J., Huang, D., Xie, M., Guo, J., Yang, J., Bao, H., Zhang, M., & Zheng, C. (2025) · Discover Education · doi:10.1007/s44217-025-00446-1 · Cited by 5
ChemLit-QA: a human evaluated dataset for chemistry RAG tasks
Wellawatte, G. P., Guo, H., Lederbauer, M., Borisova, A., Hart, M., Brucka, M., & Schwaller, P. (2025) · Machine Learning Science and Technology · doi:10.1088/2632-2153/adc2d6 · Cited by 18
Investigating How Computer Science Researchers Design Their Co-Writing Experiences With AI
Roffarello, A. M., Calò, T., Scibetta, L., & Russis, L. D. (2025) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713205 · Cited by 14
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.




