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 4,601–4,700 of 4,844
A hybrid quantitative–qualitative RAG approach for territorial well-being decision support
Comiti, G., Bisgambiglia, P.-A., & Lameta, N. (2026) · Machine Learning with Applications · doi:10.1016/j.mlwa.2026.101005
Comparative Evaluation of Advanced RAG Architectures for Medical Question Answering
Haneela, A. L. S., Sujith, R., Ganesh, A., Kudalkar, V. P., & O, P. C. (2026) · International Conference on Emerging Technologies in Computing and Communication (ETCC) · doi:10.1109/etcc69750.2026.11681946
Gibson, A. C. S., Castro, E. S. D., & Valiante, L. F. (2026) · Revista de Geopolítica · doi:10.56238/revgeov17n9-082
Characterizing Why Inference-Time Interventions Fail to Mitigate Sycophancy in LLMs
Sundaram, G., Venktesh, B., V, & Sundharakumar, K. B. (2026) · International Conference on Emerging Technologies in Computing and Communication (ETCC) · doi:10.1109/etcc69750.2026.11681976
“However, they frequently produce outputs that contradict their own previous statements or established facts— a phenomenon variously termed sycophancy [4], [5], hal-lucination [6], or self-inconsistency [7].”
Chaulagain, A., Acharya, D., Shrestha, L., Sherpa, N., & Mansur, J. (2026) · International Conference Computational Vision and Bio Inspired Computing · doi:10.1109/iccvbic71195.2026.11688311
Leonteva, A. V., Izumskaya-Kapitonova, V., & Kusina, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37867-5_24
Rong, S. (2026) · International Conference on Machine Learning and Intelligent Computing · doi:10.1145/3829441.3829500
Joshi, U. (2026) · Indian Journal of Community Medicine · doi:10.4103/ijcm.ijcm_174_26
“[12] Therefore, every AI-flagged issue must be verified against the manuscript text.”
Cao, J., Lv, H., Xue, J., Wang, S., Wang, H., Zhao, M., & Wang, W. (2026) · Cybernetics and Intelligence · doi:10.26599/cai.2026.9390020
Ferreira, J. C., & Rosa, I. (2026) · Frontiers in Digital Health · doi:10.3389/fdgth.2026.1865568 · Cited by 4
Detecting LLM Hallucinations via Anomaly Detection on Latent Representations
Malchiodi, D., Mas, L. D., Frasca, M., Rivolta, M. W., & Mazzon, I. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37976-4_10
Xu, S., Li, T., Lou, S., & Li, P. (2026) · IEEE Access · doi:10.1109/access.2026.3733278
Soni, D., Sankathala, S., & Das, S. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-2600-0383-1.ch002
Nguyen, M., & Tran, V. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-5644-0_30
A Controlled Study of Evidence Quality in LLM-Based Numerical Claim Verification
Le, Q. T., Badache, I., & Hamri, M. E. A. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-39150-6_9
Alkwai, L. M. (2026) · International Journal on Digital Libraries · doi:10.1007/s00799-026-00451-1
“Ji et al. [18] provide a comprehensive taxonomy of hallucination types, while mitigation strategies such as retrieval-augmented generation (RAG) [19] and advanced prompting techniques [20] aim to constrain model behavior.”
Sunandar, J. D. W., & Akbar, M. (2026) · Jurnal Riset Informatika · doi:10.34288/jri.v8i4.590
Automatic AI-Based Extraction of Processes from Administrative Texts: The Case of Waste Management
Membrini, A., Rossi, M. T., Mariani, L., et al. (2026) · Lecture notes in business information processing · doi:10.1007/978-3-032-37877-4_23
A safety-oriented multimodal decision framework for venipuncture assistance
Huang, Z., Xu, Y., Wang, Y., Peng, Y., Huang, X., Feng, S., Tang, X., Jiang, W., Zhang, Y., Lai, J., et al. (2026) · Pattern Recognition · doi:10.1016/j.patcog.2026.114982
Self-citationBa, A., & Mensah, P. (2026) · Journal of Public Policy and Administration · doi:10.11648/j.jppa.20261003.17
Effects of proactivity, autonomy, and transparency on user perception in human-AI agent interactions
Liu, N., & Zuo, Y. (2026) · International Journal of Industrial Ergonomics · doi:10.1016/j.ergon.2026.104058
Paul, S. (2026) · IEEE International Conference on Instrumentation (INSTCon) · doi:10.1109/instcon69741.2026.11691738
“Comparing their hallucination behavior can reveal whether hallucinations stem from scale, data, architecture, or task design, yet such analyses remain largely unexplored [12].”
Yu, C. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1924148
Attarassi, Y., & Karkouri, J. A. (2026) · Smart Cities · doi:10.3390/smartcities9090154
Ioannou, I. I., & Javaid, S. (2026) · Energies · doi:10.3390/en19184444
Li, Q., Luo, H., Peng, S., et al. (2026) · Applied Soft Computing · doi:10.1016/j.asoc.2026.116473
An AI–IoT–Cloud Platform for Smart Tea-Tourism: MQTT Sensing and LLM Services
Sun, F., Zhang, J., & Shen, X. (2026) · International Conference on Artificial Intelligence, Internet of Things and Cloud Computing Technology (AIoTC) · doi:10.1109/aiotc71250.2026.11688637
“Second, off-the-shelf LLM chatbots are known to hallucinate when answering specialized tourism questions [3] , and their free-form outputs hardly satisfy the rigid requirements of route planning or business intelligence.”
Nguyen, B. T., Pham, H. V., Tran, T. K., & Dao, S. V. T. (2026) · Intelligence-Based Medicine · doi:10.1016/j.ibmed.2026.100467
Stueckle, C. A., Limbrock, M., & Haage, P. (2026) · RöFo - Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren · doi:10.1055/a-2944-0073
Kim, S., 송용현, Kim, B., Kim, H., & Kuc, T. (2026) · Electronics · doi:10.3390/electronics15184328
Huang, L., Cui, Q., Yu, L., & Dai, J. (2026) · International Conference on Artificial Intelligence, Internet of Things and Cloud Computing Technology (AIoTC) · doi:10.1109/aiotc71250.2026.11688541 · Cited by 1
“(2) We introduce an incremental dynamic updating mechanism that assimilates newly acquired documents on the fly [6] .”
CARE: Clinically Aligned Retrieval Evidence for Consistent Radiology Report Generation
Zhu, J., Gu, Y., & Ma, J. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-38059-3_6
Zhang, S. (2026) · International Conference on Artificial Intelligence, Internet of Things and Cloud Computing Technology (AIoTC) · doi:10.1109/aiotc71250.2026.11688830
“Ji et al. [15] and Huang et al. [16] pointed out that in high-noise or information-poor conditions, LLMs are highly prone to generating "hallucinations", resulting in homogenized outputs and factual errors.”
Brena-Melendez, A., Avina-Bravo, E. G., Welti‐Chanes, J., Alfaro-Ponce, M., & Tejada-Ortigoza, V. A. (2026) · LWT · doi:10.1016/j.lwt.2026.119965
Encoder Language Models for Zero-Shot Recommender Systems: Cross-Domain and Cross-Lingual Evaluation
Müller, P., Walek, B., & Farana, R. (2026) · Computers · doi:10.3390/computers15090639
Ni, L., Pei, J., Wang, H., & Chen, H. (2026) · Electronics · doi:10.3390/electronics15194348
Yu, S.-Y., & Du, J.-S. (2026) · Agricultural Research · doi:10.1007/s40003-026-01053-8
Prabha, S., Collaço, B. G., Malena, C. R., Haider, S. A., Wood, N. G., & Forte, A. J. (2026) · Bioengineering · doi:10.3390/bioengineering13101097
PathLit-Agent: A Multi-stage AI Literature Mining System for Pathway-Focused Biomedical Insight
Maghsoudi, Z., & Harris, F. C. (2026) · Advances in intelligent systems and computing · doi:10.1007/978-3-032-27214-0_35
Fusion of fuzzy rule-based factual and counterfactual explanations with large language models
Perez-Ferreiro, P. M., Alonso, J. M., Catalá, A., & Bugarín, A. (2026) · International Journal of Approximate Reasoning · doi:10.1016/j.ijar.2026.109828 · Cited by 1
Zhou, C., Wu, H., Li, G., & CHEN, Y. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1915597
“However, recent evaluations indicate that conversational fluency does not necessarily guarantee reliability, factual accuracy, or appropriate support, particularly in health-related contexts (Ji et al., 2023; Langston et al., 2024; Stade et al., 2024).”
COMPARATIVE ANALYSIS OF ENGINEERING DOCUMENTATION USING DETERMINISTIC AND LLM-BASED METHODS
Myhovych, A. V., & Kunytsia, A. (2026) · Journal of Rocket-Space Technology · doi:10.15421/452571
Kholmurotov, R. (2026) · Research Square · doi:10.21203/rs.3.rs-11114152/v1
Xing, Y., Zhang, Q., Li, S., Li, K., Yao, Z., Zhao, J., & Wang, Z. (2026) · Electronics · doi:10.3390/electronics15194360
Liang, Z., Liu, L., & Peng, T. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7507204
Yang, S., Xu, Y., Jia, C., & Wang, Y. (2026) · Information Processing & Management · doi:10.1016/j.ipm.2026.105179
Self-citationTripathi, C., Kumari, S., Bobde, T., Chatur, S., Dhomane, V., & Pethe, S. (2026) · 2026 International Conference on Computing, Intelligence, and Applications (CIACON) · doi:10.1109/ciacon70148.2026.11689445
Augmented Reality for STEM Education: A Scalable Model for Remote Classrooms
Vandana, Kaur, R., Quraishi, S., et al. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-28100-5_19
AI Agents and Agentic AI in plant breeding: new frontiers, opportunities and challenges
Vitale, P., Acharya, D. B., Gerard, G., Dreisigacker, S., Montesinos-López, O. A., Montesinos‐López, A., Ortiz, R., & Crossa, J. L. (2026) · Frontiers in Genetics · doi:10.3389/fgene.2026.1958764
“In plant breeding, hallucinations may manifest as fabricated model results, incorrect interpretations of trial analyses, invented parameter values, misdescriptions of experimental designs, or unsupported recommendations for line advancement or crossing (Ji et al., 2023).”
KG-MAFMEA: A Knowledge Graph-Driven Multi-Agent Framework for Failure Mode and Effects Analysis
Wang, S., Yang, Y., Gan, W., & Hu, H. (2026) · International Conference on System Reliability and Safety Engineering (SRSE) · doi:10.1109/srse70977.2026.11694991
“However, direct use of LLMs in reliability and safety engineering may produce plausible but unsupported conclusions [10].”
Vantaku, C., & Avakian, A. (2026) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-026-02326-y
Wang, K., Pan, Z., & Lu, Y.-B. (2026) · Computers in Human Behavior · doi:10.1016/j.chb.2026.109201
Ravikumar, S. (2026) · Information Processing & Management · doi:10.1016/j.ipm.2026.105174
Kabli, A. M. (2026) · Digital Health · doi:10.1177/20552076261492994
Clever But Not Wise: The Limits of AI and the Space of the Human
Wilby, M. (2026) · Human Affairs · doi:10.1515/humaff-2026-0023
Li, T. (2026) · International Conference on Electronic Information, Computer and Aerospace Remote Sensing (EICARS) · doi:10.1109/eicars71536.2026.11694433
Personalizing Older Adults’ Cognitive Training Robot Games with LLM Support
Canapa, G., Catricalà, B., Manca, M., Paternò, F., Santoro, C., & Zedda, E. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2731297
Gautam, G., Adhikari, P., Sah, N., Thapa, R., Cheol, L., & Shrestha, B. (2026) · IEEE Access · doi:10.1109/access.2026.3733686
“Ji et al. [1] categorized LLM hallucinations as intrinsic where the output contradicts the source, and extrinsic, where it cannot be verified from the source.”
Dergaa, I., Othman, R. B., Ghouili, H., Ceylan, H. İ., Souissi, N., Bchini, S., Ștefănică, V., Muntean, R. I., & Bragazzi, N. L. (2026) · Nutrients · doi:10.3390/nu18193128
Gonçalves, K. P. F. (2026) · Nexus Science Review · doi:10.66104/027mff80
Li, H., Shi, J., Chen, X., Jia, Q., Wang, R., Peng, Y., Gao, L., Chen, Q., An, H., Ouyang, R., et al. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-38470-6_41
Rustamov, Z., Pashayev, M. A., Jafarov, S., Aslanli, U., Mammadov, A., Valizada, A., & Rustamov, S. (2026) · Journal of Advances in Information Technology · doi:10.12720/jait.17.9.1716-1723
“Prior work has documented risks related to hallucination, inconsistent behavior, and lack of transparency in generative models, particularly when applied to high-stakes decision-making contexts [6, 17].”
Enterprise Conversational AI with Retrieval-Augmented Generation: A CRM Case Study
Sasinka, M., Kotyrba, M., Volná, E., & Pavlíček, M. (2026) · Acta Informatica Pragensia · doi:10.18267/j.aip.320
EmERGE: A Multimodal LLM-Based Evidence-Mediated ECG Report Generation and Evaluation Approach
Yao, K., Liu, Y., Wang, X., & Shen, G. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-38470-6_44
Al-Saqaf, W. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-9190-8.ch018
Bian, Q., Xin, L., & Ma, B. (2026) · Public Administration · doi:10.1111/padm.70097
Acharya, N., & Kar, P. (2026) · Frontiers in Future Transportation · doi:10.3389/ffutr.2026.1951756
Comparative evaluation of ambient digital scribe systems in clinical documentation
Weiss, E., Zsenits, B., Nasar, F., Phillips, M., Mudgundi, V., Salhab, T., Muth, A., Zhu, K., Kolosko, J., Hansen, S., et al. (2026) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocag157
Explanation Transparency and Trust Perception After Generative AI Failure: A Dual-Path Model
Yang, Q.-Y., & Dong, H.-L. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2734323
Wendakoon, K., & Tripathi, N. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-39766-9_11
Balancing Engagement and Polarization: Multi-Objective Alignment of News Content Using LLMsPreprint
Cheng, M., Ofek, E., & Yoganarasimhan, H. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5226629 · Cited by 3
Navigating AI-Accelerated Science: An Intellectual Virtue Approach to Responsible Inquiry
Sikimić, V. (2026) · Episteme · doi:10.1017/epi.2026.10140
Situational Context Embedding and Knowledge-Based Grounding for LLMs in Surgical Training Assistance
Malayeri, A. B., Seibold, M., Oettl, F. C., Cavalcanti, N., Gerth, T., Lekar, R., Hoch, A., Zingg, P., & Fürnstahl, P. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-38233-7_47
SAS Multimetric Tribunal: Structural Classification of Hallucination in AI-Generated TextPreprint
Durante, G. E. (2026) · Qeios · doi:10.32388/t420iv.3
Sapiński, A. (2026) · Academic Visions · doi:10.66556/2786-586x.54.sapiski-a
Chokki, A. P., Njike, T. N., & Daune, J.-F. (2026) · Procedia Computer Science · doi:10.1016/j.procs.2026.08.467
Yertay, A., Bakirov, K., Shoman, A., Kudashov, Y., Baktash, A. P., Tussupov, J., & Yenlik, A. (2026) · Procedia Computer Science · doi:10.1016/j.procs.2026.08.367
Stephan, D., Schumacher, S., Al‐Nawas, B., Kämmerer, P. W., & Thiem, D. G. E. (2026) · Oral and Maxillofacial Surgery · doi:10.1007/s10006-026-01643-4
“This robustness addresses one of the key concerns associated with LLMs, namely the risk of “hallucinations” or fabricated content [22 ], and suggests that, under defined and supervised conditions, AI can operate dependably in bounded linguistic transformation tasks such as simplifica - tion.”
Mais, R. E., Hamze, N., Abbas, C., Mubasher, H. A., & Awad, M. (2026) · Annals of Computer Science and Information Systems · doi:10.15439/2026f3733
A LLM-Powered Multi-agent System for Cross-View Integration in Digital Government Systems
Xie, C., Shang, H., Li, C., & Huang, K. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-5699-0_41
Chunk Size and Retrieval Depth Optimization in a Minimal Retrieval-Augmented Generation Pipeline
Ahmad, S., Hussain, J., Najeeb, H., Ahmad, F., Shah, S. H., & Ibad, M. S. (2026) · International Journal of Innovations in Science and Technology · doi:10.33411/ijist/2053
kuppani, C., & Halavar, B. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-38062-3_51
IteraTri: Detecting Hallucination of LLM in Knowledge-Intensive QA via Iterative Verification
Wang, Y., Zhang, J., Yu, D., & Li, X. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-5693-8_24
Hwang, W., Nurtantyana, R., & Lin, T.-W. (2026) · Interactive Learning Environments · doi:10.1080/10494820.2026.2735914
Ghaywat, V., Singh, A., Shahade, A. K., Khan, W., & Deshmukh, P. V. (2026) · Discover Artificial Intelligence · doi:10.1007/s44163-026-02373-y
Beyond Textual Accuracy: Evaluating RAG-Based Assistant for Novice Operators in Industrial Robotics
Botticella, E., Wu, Z., Aliev, K., & Bruno, G. (2026) · IFIP advances in information and communication technology · doi:10.1007/978-3-032-38618-2_7
Learning with RAG-Enhanced Language Models: A Mixed-Methods Study
Miladi, F., Psyché, V., Diattara, A., Mawas, N. E., & Lemire, D. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-39106-3_12
A pragmatic workflow for integrating large language models into toxicological risk assessment
Bassan, A., Amberg, A., Barreca, M. L., Bradley, P. M., Bringezu, F., Souza, J. D. P., Guicheney, C., Piparo, E. L., Kovarich, S., Massarsky, A., et al. (2026) · Frontiers in Toxicology · doi:10.3389/ftox.2026.1877128
A RAG- and Blockchain-Based System for Souvenir Ticket Evidence Preservation and Traceability
Chen, H.-C., Li, H.-Y., Hsu, P.-Y., et al. (2026) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-032-38477-5_22
Correa-Jaramillo, C., & Vásquez, J. C. G. (2026) · Sustainability · doi:10.3390/su18199910
Sztuczna inteligencja w warsztacie edukatora-popularyzatora. Propozycja dydaktyczna
Gałuszka, J. (2026) · Archeion · doi:10.4467/26581264arc.26.005.23512
Hybrid Sparse-Dense RAG for Citation-Grounded Legal QA Over SCOTUS Case Law
Sherif, A., Sharaf, A., & Adel, S. (2026) · Intelligent Methods, Systems, and Applications (IMSA) · doi:10.1109/imsa70415.2026.11700297
“LLMs demonstrate strong question-answering capabilities [2], [3], but direct application to legal tasks is problematic: parametric knowledge is frozen at training time and models can hallucinate plausible-sounding but unsupported legal statements [4].”
Pramudya, A., Haryanto, V. C. P., Ramli, N. V., Saputra, C. H., Hartono, S., & Reyes, I. V. P. D. (2026) · International Conference on Information Management and Technology (ICIMTech) · doi:10.1109/icimtech202671078.2026.11698921
“In the case of extrinsic hallucination, the generated information cannot be verified in any external source [21].”
Analysis of Generalist and Domain-Specific Large Language Models for Medical Assessment
Maheswara, I. G. A., Yudistira, F. S., Prasetyo, A., & Majiid, M. R. N. (2026) · International Conference on Information Management and Technology (ICIMTech) · doi:10.1109/icimtech202671078.2026.11699040
“Hallucination in the medical scope can have a very threatening impact on patient safety [1], [4], [9].”
Shukla, U., Jaishetty, H., Anjali, R., & Anand Kumar, M. (2026) · IEEE Region 10 Symposium (TENSYMP) · doi:10.1109/tensymp68909.2026.11704054
“Prior hallucination detection methods rely on object detectors [10] or vision-language similarity metrics [5] applied post-hoc; hallucination in LLMs and VLMs has been extensively studied [8], [9] but T2I failures have distinct geometric causes orthogonal to text rendering bias [7].”
Selective Extraction: A Hallucination-Free Approach to LLM-Based Document Information Extraction
Rao, V., Saraogi, Y., Reddy, T., & Purohit, M. V. (2026) · IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT) · doi:10.1109/conecct68978.2026.11700668
“Hallucination in LLMs refers to the generation of content that is plausible but factually incorrect, very different from the source material, or entirely fabricated [3], [9].”
Beyond Context Window: Dynamic Knowledge Graphs for Lifelong Learning Llms
Desai, V., Kalshetty, R., M, M. S., & Kotagi, S. (2026) · IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT) · doi:10.1109/conecct68978.2026.11700592
“…gains demonstrated in Table II (F1 = 0.89 vs. 0.76 for RAG) are notable in the context of LLM hallucination—a systemic challenge in knowledge-intensive generation tasks [24]—suggesting that externalized structured memory can help mitigate factual inconsistency in long-horizon conversational AI.”
InfluentialPai, R., Vineeth, K. H. M., Shetty, C., Nayak, G., & Aditya, C. (2026) · IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT) · doi:10.1109/conecct68978.2026.11700597
“Second, hallucination phenomena—wherein models generate syntactically valid but semantically incorrect SQL referencing non-existent tables or columns—remain a persistent challenge [18].”
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.




