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 3,501–3,600 of 4,838
PROMPT ENGINEERING IN MARKETING AND ECONOMICS: METHODOLOGY AND METHODS
Сидорчук, Р. (2025) · Drukerovskij vestnik · doi:10.17213/2312-6469-2026-2-371-389
The Role of Prompt Engineering in Reducing Hallucination in Generative AI Models
Subbarao, M. V., Kalyani, S., Shreya, S., et al. (2026) · International Conference on Emerging Smart Computing and Informatics (ESCI) · doi:10.1109/esci68015.2026.11493253
“Prompts with thoughtful, design, including zero-shot, few-shot, role-based, or chain-of-thought forms, can aid LLMs in comprehending user intent and producing factually consistent answers [8].”
Lee, D., Choi, J., Yi, G. H., Sohn, S. S., Lee, B., & Kim, D.-H. (2026) · Digital Discovery · doi:10.1039/d6dd00028b · Cited by 1
Brooks, B. A. (2026) · Nurse Leader · doi:10.1016/j.mnl.2026.102768
Santhosh, V. N., Vas, R., Roychowdhury, B., Sakthi, K., & Rahaman, M. (2026) · Research Evaluation · doi:10.1093/reseval/rvag024 · Cited by 1
CC-RAG: A Collaborative-Critical Dual-Agent Framework for Enhanced Retrieval-Augmented Generation
Fang, Y., Cheng, Y., & Zhang, W. (2025) · International Conference on Intelligent Education and Intelligent Research (IEIR) · doi:10.1109/ieir67391.2025.11491388
“Large Language Models (LLMs) have achieved revolutionary progress in multiple domains of natural language processing, yet their application remains constrained by the inherent challenges of ”knowledge cutoff” and ”factual hallucination” [1].”
A pragmatic study of teacher dialogues with a large language model-based AI chatbot
Tan, S. C., Tan, Y. Y., Yuan, G., Teo, C. L., & Dai, L. (2026) · Linguistics and Education · doi:10.1016/j.linged.2026.101541
SHIFT: Smoothing Hallucinations by Information Flow Tuning for Multimodal Large Language Models
Wang, S., Zhang, Y., Zhu, Y., liu, E., Li, J., Liu, Y., & Ji, X. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.00347 · Cited by 5
“The hallucination phenomenon has been widely observed in the field of language models [21, 24, 30, 40, 53], where researchers find that these models often produce responses with factual inaccuracies or not aligned with the input prompt, referred as factual and faithful hallucinations, respectively.”
LLM-based Enhanced Question Answering System in Medical Domain
Ganar, R., Dharme, O. A., Gharat, S. A., Yadav, A., Parab, A. A., & Radke, M. A. (2025) · Artificial Intelligence and Cloud Computing Conference (AICCC) · doi:10.1145/3789982.3790002
Gao, Z., & Liu, Y. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.116003
Duan, Z., Zhao, G., Li, C., & Liu, B. (2026) · Biology · doi:10.3390/biology15090708
“General-purpose LLMs are inherently highly non-deterministic, often producing fluctuating responses to the same prompts and lacking an intrinsic mathematical reasoning engine [23].”
Epistemic laundering: generative AI and the naturalization of misrecognition
Kalaitzidis, T. J. (2026) · AI & Society · doi:10.1007/s00146-026-03068-9 · Cited by 2
Copyright, Privacy, and Ethical Issues of Generative AI
Thoker, Z. A., Marwaha, L., & Jan, S. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7852-7.ch003 · Cited by 2
Yang, M. (2026) · International Conference on Informatics Education and Computer Technology Applications (IECA) · doi:10.1145/3802133.3802267
“Despite their power, LLMs are prone to “hallucination,” or the generation of factually incorrect or nonsensical information, which can be dangerous in medical contexts [15] [16].”
A review of the applications and tasks of large language models in the legal field
Peng, G., Li, D., Wang, G., Guo, S., Su, L., & Yu, T. (2026) · PeerJ Computer Science · doi:10.7717/peerj-cs.3836 · Cited by 1
Self-citationMasoumi, N., Davar, O., & Eftekhari, M. (2026) · Knowledge and Information Systems · doi:10.1007/s10115-026-02778-2 · Cited by 1
Vijayakumar, K., Ambike, M., & Saraswat, A. (2026) · Discover Applied Sciences · doi:10.1007/s42452-026-08787-8
“Hallucina-tions generate highly plausible but completely fabricated information that passes human inspection [11].”
Mitigating Intrinsic Hallucinations Using Orthogonal Fine-Tuning and Dirichlet Prior Calibration
Kaushal, S., Singh, C., Bijalwan, V., & Mohana, R. (2026) · World Conference on Computational Science and Technology (WcCST) · doi:10.1109/wccst67302.2026.11495718
“Hallucination occurs when a model produces a confident but factually incorrect statement, creating an output that appears coherent and plausible while diverging from verifiable reality [4].”
Vibe Econometrics and the Analysis ContractPreprint
Ashton, L. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6699999 · Cited by 2
Simple picture of how output from ChatGPT-like AI shifts from good to bad
Johnson, N. F., & Huo, F. Y. (2026) · PNAS Nexus · doi:10.1093/pnasnexus/pgag148 · Cited by 3
Generative AI for Financial Translation: Opportunities and Challenges in Hong Kong
Siu, S. C. (2026) · New frontiers in translation studies · doi:10.1007/978-981-95-4374-8_6
End-To-End LLM Systems: Data, Training, Evaluation, and Deployment
Dhanasekar, S. (2026) · International Conference on Machine Learning and Autonomous Systems (ICMLAS) · doi:10.1109/icmlas67792.2026.11483701
“The fast development of AI has moved beyond simplistic statistical learning to the world of large neural networks that are able to reason like humans, and that change has been recorded through the historical research of AI development by computational intelligence researchers [6].”
MRE-Guard: A RAG-Based Multi-Agent Framework for Phishing Detection
Chang, X., & Shi, K. (2026) · International Conference on Electronics Technology and Artificial Intelligence (ETAI) · doi:10.1109/etai68332.2026.11485151
“First, LLMs are prone to hallucination [6], potentially generating confident but incorrect judgments, leading to high false positive rates.”
GenAI Inside: A Practical Methodology for Embedding AI Across Enterprise Workflows
Aliyu, M., & Barros, J. (2025) · Artificial Intelligence for Business (AIxB) · doi:10.1109/aixb65684.2025.00022
“Recent evaluation initiatives [19], [20] Fig.”
A Method for Data Structure Question Generation Using Enhanced Large Language Models
Ma, D. (2026) · International Conference on Electronics Technology and Artificial Intelligence (ETAI) · doi:10.1109/etai68332.2026.11485367
“…LLMs to question generation presents several limitations: (1) When the model’s intrinsic knowledge is insufficient, it is prone to “hallucinations” [5], making it difficult to guarantee content correctness; (2) Inadequate Outline Adherence: The model’s insufficient grasp of specific exam outlines…”
Multi-modal RAG for Text and Image Retrieval
Nidhi, M., Rao, D., Vernekar, S. M., Itigowni, P., Kulkarni, U., & Narayan, D. G. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-19185-4_26
Quality Evaluation of Software Functional Requirements Generated by LLMs: A Systematic Mapping Study
Lourenço, L. M., Bezerra, C. I. M., & Marques, A. P. (2026) · Anais do Simpósio Brasileiro de Sistemas de Informação (SBSI) · doi:10.5753/sbsi.2026.248317
“However, the adoption of LLMs in this context still faces important challenges, such as hallucinations [Ji et al. 2023], excessive dependence on prompts, lack of domain knowledge, and difficulty in ensuring semantic correctness and alignment with stakeholders’ needs.”
Yoon, S. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2657559
Kurra, S. K., Ruksana, S., & Borusu, V. (2026) · IEEE International Conference on Emerging Computing and Intelligent Technologies (ICoECIT) · doi:10.1109/icoecit68303.2026.11496690
Text Message Routing System for Chat-Based Applications
Angelo, B. U. D., Zanetti, G. G., Souza, A. F. D., Badue, C. S., Jacobsen, A. G., & Oliveira-Santos, T. (2026) · Anais do Simpósio Brasileiro de Sistemas de Informação (SBSI) · doi:10.5753/sbsi.2026.248585
“Hallucinations in generative language models are still an open problem [Ji et al. 2023], and occurs when models generate responses that seem to be accurate to a non-expert but contain erroneous or misleading information.”
Next - Gen University Helpdesk (NexU)
S.Hemaswathi, TamilMuthu, K., Hariharasuthan, K. B., Sathiya Seelan, R. P., & Sarveen Bala, R. (2026) · IEEE International Conference on Emerging Computing and Intelligent Technologies (ICoECIT) · doi:10.1109/icoecit68303.2026.11497274
“Finally, NexU works as a comprehensive academic assistant that underpins the digital transformation efforts in education by improving the levels of accessibility, automation, and personalization [4], [22].”
Supporting Reviewing Reviews: How HCI Authors Handle Peer Reviews of Manuscripts
Yeung, C. A., Stark, J., Li, J., Chevalier, F., Park, J., Kim, Y., & Tang, A. (2025) · International Conference on Human-Engaged Computing (ICHEC) · doi:10.1145/3786995.3786998
Turjya, S. M., Sarker, O., Chayon, D. C. S., et al. (2025) · International Conference on Computer and Information Technology (ICCIT) · doi:10.1109/iccit68739.2025.11489538
“Nevertheless, these models are usually plagued by two intrinsic challenges: factual hallucination and limited generalizability across domains [3].”
A visualization framework for hierarchical structure enhanced mind mapping
Liu, Y., Wang, H., Lu, W., Chen, W., & Ji, Y. (2026) · Visual Informatics · doi:10.1016/j.visinf.2026.100325
Examining the Influence of ChatGPT on Programming Courses in Higher Education
Prado, J., & Palaoag, T. D. (2026) · International Conference on Advances in Artificial Intelligence and Machine Learning (AAIML) · doi:10.1109/aaiml67890.2026.11498114
“More broadly, technical reports and failure taxonomies on LLMs document recurring limitations, including hallucinations, reasoning errors, and inconsistent behavior, reinforcing the view that prompt-dependent outputs must be validated rather than treated as authoritative solutions [10][11].”
ChatGPT in Start-up Consulting
Metzner, N., & Miller, A. (2025) · Zeitschrift für Klein- und Mittelunternehmen/Zeitschrift für KMU und Entrepreneurship · doi:10.3790/zfke.2025.1479404
Enhancing Security Requirements Coverage via RAG and Automated Feedback Loops
Sabetta, G., Cannavale, A., Lucia, A. D., & Palomba, F. (2026) · IEEE International Conference on Software Analysis, Evolution and Reengineering - Companion (SANER-C) · doi:10.1109/saner-c67878.2026.00045
“This approach has demonstrated effectiveness in mitigating factual errors and reducing the generation of ungrounded content [13].”
A survey of large language models: techniques, applications, and challenges
Ji, J., Zhao, X., & Shi, X. (2026) · Multimedia Systems · doi:10.1007/s00530-026-02288-9
“Financial applications exhibit similar disparities: Ji et al. (2023) demonstrated that credit scoring models trained on historical loan data rejected applications from low-income neighborhoods at a 28% higher rate, despite controlling for creditworthiness metrics, due to spurious correlations…”
Rajakarunanayake, S., & Weerasinghe, B. A. I. R. (2026) · International Research Conference on Smart Computing and Systems Engineering (SCSE) · doi:10.1109/scse70081.2026.11499873
Barla, N. S., & de Berardinis, J. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-25156-5_25
“LLMs are prone to factual invention and distortion – a phenomenon widely known as “hallucination” – which renders them fundamentally unsuitable for domains where factuality is paramount [20, 36].”
Chen, D., Gong, Y., Chen, J., & He, R. (2026) · Energy and Buildings · doi:10.1016/j.enbuild.2026.117578 · Cited by 3
Not All Language Models Are Generative: A Methods Primer on Encoder Models in Social Work ResearchPreprint
Perron, B. E., Stoll, D., Qi, Z., et al. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6731299
Paul, S., Mitra, A., Podder, A., & Roy, P. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7320-1.ch008
Lu, Y., & Jia, C. (2026) · Computer Assisted Language Learning · doi:10.1080/09588221.2026.2665351 · Cited by 1
Wei, S., Lai, Z., Fu, B., & Shi, X. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2026.3689656
“…some studies indicate that the intrinsic knowledge obtained from LLMs training may be incomplete or outdated, which can lead to hallucination problems, such as generating statements with incorrect facts that may lead to erroneous stance analysis of the target in stance detection tasks [17], [18].”
Energy Domain Ontology Matching with Large Language Model
Pan, Z., Li, H., & Monti, A. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-25159-6_18
Afeku-Amenyo, H., & Archibong, E. I. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7320-1.ch004
Method Distinction for Hallucination Detection
Gupta, B. P., Gayratovich, E. N., & Madaminov, B. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7320-1.ch007
Generative AI in Malaysian Pharmacy Education: A Call for Responsible Integration
Chang, C. T., Wong, Y. J., Choong, T. C., Cheong, W. L., & Chan, H. (2025) · Malaysian Journal of Pharmacy · doi:10.52494/maljpharmv11202
Explainable ICD coding via MediThoth: A multimodal clinical decision support system
Elbehery, M. M., Mohamed, M. S., & Elshafey, M. A. (2026) · Array · doi:10.1016/j.array.2026.100875
LogicJitter: Let LLMs play and uncover misinformation
Herranz-Celotti, L., & Viviani, M. (2026) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2026.111215 · Cited by 2
Erdem, İ. A. (2026) · World Tax Journal · doi:10.59403/f1zv8g
Das, S., Ibrahim, R., Davis, N. A., Ge, Y., Guo, Y., Sarker, A., & Rowh, M. A. W. (2026) · Research Square · doi:10.21203/rs.3.rs-9254050/v1
Hallucination-Aware, Injection-Resistant LLMs for Scientific Text Integrity
Bhardwaj, N., Bhardwaj, A., & Garg, L. (2025) · International Conference on Sustainable Computing and Smart Systems (ICSCSS) · doi:10.1109/icscss64956.2025.11501150
“One of the most documented and concerning issues in generative model output is hallucination [7].”
Document-Level Relation Extraction with Bayesian-Adaptive Monte Carlo Tree Search
Wang, Z., Wang, Y., & Zhao, W. H. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-0378-9_6
Physics-informed multimodal large language models for intelligent energy question answering
Abbas, S., Amin, W., Yuyu, Z., Soleimani, A., Pinnarelli, A., Vizza, P., & Heidari, M. (2026) · Results in Engineering · doi:10.1016/j.rineng.2026.110718
Uncertainty-Calibrated Trust Modelling for LLM-Generated Misinformation Detection
Godwin, M. P., Sankar, H., Kavitha, R., Kamalesh, M., Raghunath, S., & Baig, N. I. (2026) · International Conference on Data Science, Agents & Artificial Intelligence (ICDSAAI) · doi:10.1109/icdsaai69492.2026.11505140
“LLMs are highly susceptible to “hallucinations”—generating factually incorrect yet structurally persuasive content—which poses substantial risks when deployed in domains such as automated journalism, medical summarisation, and legal analysis [3].”
Multi-Route Search and Adaptive Fusion for Power QA with Small Language Model Guidance
Shen, Z., Guo, Q., Jiang, L., Huang, J., Yu, Z., Cai, X., Pang, H., & Yu, T. (2026) · Algorithms · doi:10.3390/a19050378 · Cited by 2
Self-citationCoT-F: Leveraging Chain-of-Thought Families in Large Language Models for Complex Question Answering
Zhao, F., Liu, X., Wang, L., Zhao, R., Yang, Y., & Xu, G. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-0369-7_35
Consensus-on-Graph: Plan-Driven Exploration and Consensus Decision-Making on Knowledge Graphs
Hu, C., Zhan, B., Yuan, L., Fan, W., & Yan, S. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-0369-7_32
FAST-RRG: Efficient and Effective Retrieval-Augmented Generation for Chest X-Ray Report Generation
Cai, Y. (2026) · International Conference on Communications, Information System and Computer Engineering (CISCE) · doi:10.1109/cisce69494.2026.11504480
“While Multimodal Large Language Models (MLLMs) [2], [3] generate fluent reports, their clinical deployment is constrained by hallucinations, posing unacceptable risks in healthcare environments [4].”
Apshinashvili, I., & Pkhakadze, G. (2026) · Georgian Medical Journal · doi:10.66636/gmj.v1.i2.a113
“Third, opacity — when a model states that a symptom pattern “is consistent with” a named condition, the underlying reasoning is not traceable to a specific clinical guideline, making independent review by a second clinician impossible [6,7].”
Zhao, R., Zhao, F., & Xu, G. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-0369-7_34
LLM-Enabled Cognitive Search for Context-Aware Knowledge Retrieval from Heterogeneous Data Sources
Gouru, M., Mittal, T., Syed, A., Prasad, S. K., Bolla, S. R., & Bandaru, R. R. (2026) · IEEE International Conference for Convergence in Computing Technology (I3CTCON) · doi:10.1109/i3ctcon68242.2026.11507233
“Standalone LLMs, however, have problems of hallucination, insufficiently based on enterprise data, and inefficiencies in operating with large external bodies of knowledge [8].”
Leffrang, D., Passlack, N., Müller, O., & Posegga, O. (2026) · Business & Information Systems Engineering · doi:10.1007/s12599-026-01000-5 · Cited by 3
Diagnostic accuracy and citation integrity of four large language models on otolaryngology vignettes
Pennington-FitzGerald, W., Warrier, A., Durant, S., Sharaf, I., Carlino, F., Pamula, S. D., Eloy, J. A., & Levi, J. (2026) · European Archives of Oto-Rhino-Laryngology · doi:10.1007/s00405-026-10253-5
“This aligns with recent findings showing that even high-performing LLMs, including those with RAG capabilities, frequently cite unverifiable or misattributed references [13, 27–30 ].”
Trust in Large Language Models: An Exploratory Validation Framework
Money, W. H., & Mew, L. (2026) · Journal of Information Systems Applied Research and Analytics · doi:10.62273/npbz2264
Hierarchical Knowledge Graph-Augmented Large Language Model for Learning Path Generation
Hu, T., & Hao, G. (2026) · International Conference on Artificial Intelligence, Digital Media Technology and Social Computing (ICAIDS) · doi:10.1145/3806262.3806285
Wang, K., Wang, T., Wang, B., Li, Q., Liu, Y., Han, Q., Zhang, C. L., & Ren, Y. (2026) · Reviews in Fish Biology and Fisheries · doi:10.1007/s11160-026-10058-z · Cited by 2
“Such inac-curacies often arise because LLMs generate coherent text from learned statistical patterns rather than verified information (Ji et al. 2023).”
Janani, S., & Das, A. (2026) · International Conference on Inventive Computation Technologies (ICICT) · doi:10.1109/icict68280.2026.11510860
“Ji et al. [1] conducted a comprehensive survey that classified hallucinations into intrinsic and extrinsic categories, emphasizing that even sophisticated language models frequently produce fluent yet factually inaccurate content.”
Grounding large language models in hydrologic modelling
Li, Y., Hu, S., Zhang, F., Yu, X., Luo, W., Liu, D., Xu, B., & Zhao, J. (2026) · Journal of Hydrology · doi:10.1016/j.jhydrol.2026.136358 · Cited by 1
Yin, B., & Lu, S. (2026) · International Conference on Computer Research and Development · doi:10.1145/3796263.3796286
Morphodynamic Meta-Cognition: A Framework for Self-Regulating Intelligence SystemsPreprint
Mavroudis, I., Ciobica, A. S., Gurzu, B., Anton, E. T.-T., & Novac, B. (2026) · Research Square · doi:10.21203/rs.3.rs-9641646/v1
Dai, A. X. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2667475
Artificial Intelligence and Digital Forensics: How They Support Each Other
Thuraisingham, B., & Akbar, K. A. (2026) · Signals and communication technology · doi:10.1007/978-3-031-98036-7_10
AGWAN, S., Tambe, A., Shah, S., & Pole, G. S. (2026) · Journal of Ambient Intelligence and Humanized Computing · doi:10.1007/s12652-026-05091-6
“However, such systems also introduce critical risks, including hal-lucinated facts (Ji et al. 2023), harmful or biased outputs (Gehman et al. 2020), Ganguli et al. (2023), and inconsistent behavior across user personas and contexts (Xu et al. methods use powerful models to evaluate responses…”
InfluentialKI-Chatbots: Was über ihre Akzeptanz entscheidet
Alyouzbashi, M. K., Dopler, G., & Ebner, R. (2026) · Wissensmanagement · doi:10.1007/s43443-026-0716-8
de Carvalho, L. M., Peças, P., & Jorge, D. P. (2026) · Sustainability · doi:10.3390/su18104787 · Cited by 1
Wang, S., Su, J., Han, X., Kang, D., Zhang, X., Xu, F., Liu, C., Pan, J., Wang, X., Zhan, Q., et al. (2026) · Advanced Science · doi:10.1002/advs.75709
“Directly employing LLMs to parse high-resolution MPM images for generating diagnostic descriptions poses a fundamental challenge, as these models struggle to extract diagnostically meaningful global semantic features directly from pixels [ 48 ].”
When it Comes to Scientific Information Extraction and LLMS, Less is More
Shaik, S., Blaiszik, B. J., & Tchoua, R. (2026) · International Journal of Semantic Computing · doi:10.1142/s1793351x26420018
Stepanov, N., Radaeva, A., Panfilov, P., Suleykin, A., & Pyatetsky, V. (2026) · Big Data and Cognitive Computing · doi:10.3390/bdcc10050156
“Using large language models (LLMs) without grounding in primary sources increases the risk of unverifiable claims and inconsistent responses [1].”
SES-RAG: An Adaptive Framework with ModernBERT Segmentation and Generative-Alignment Expansion
Gao, C. S., He, H., & Zhang, Q. (2026) · Applied Sciences · doi:10.3390/app16104849
Liu, M. (2026) · International Conference on Computer, Artificial Intelligence and Control Engineering (CAICE) · doi:10.1145/3804601.3804625
“This study also follows this approach, but focuses on textual narratives rich in cultural sym-bolism, while using retrieval enhancement paradigms to improve fidelity and traceability [4-5].”
Zheng, X., Chen, T. X., & Zhang, W. (2026) · International Journal of Cognitive Computing in Engineering · doi:10.1016/j.ijcce.2026.05.001
Human–Machine Interaction and AI Coaching
Olafsson, B., Passmore, J., & Terblanche, N. H. D. (2026) · Wiley eBooks · doi:10.1002/9781394292806.ch22
Esina, S. V., & Inkpen, D. Z. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-20322-9_27
Multimodal Fusion with External Knowledge (MFEK)
Lin, Z., & Li, Q. (2026) · Wireless networks · doi:10.1007/978-3-032-20416-5_3
Face value: How avatar identity shapes epistemic trust in AI-mediated learning
Anthis, Z., & Kyriakidou-Zacharoudiou, A. (2026) · Computers and Education Artificial Intelligence · doi:10.1016/j.caeai.2026.100610 · Cited by 1
Exploring the Use of Generative AI for Assessing Data-Driven Stories
Duca, A. L., & Yocco, V. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-16448-3_13
Auditing Hallucination in Lightweight Vision-Language Models for Radiology Captioning
Wang, Z., Zhang, H., Xie, C., et al. (2026) · International Conference on Generative Artificial Intelligence and Information Security (GAIIS) · doi:10.1109/gaiis69281.2026.11519346
“Hallucination, broadly defined as output content not faithfully grounded in the input or reference evidence, remains a persistent failure mode in generative models [5], [6].”
Zhang, S., Li, H., Jiang, Z., Sang, M., Liu, J., & Leong Hou, U. (2026) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2026.3688663 · Cited by 1
“Knowledge graph fact checking has been proven e ff ective for mitigating such factual hallucinations [56], [57].”
Learning to Debate: Optimal Stopping Strategies for Multi-Agent LLM Deliberation
Zhang, X., & Hu, D. (2026) · International Conference on Generative Artificial Intelligence and Information Security (GAIIS) · doi:10.1109/gaiis69281.2026.11519273
“However, single-model inference often suffers from hallucinations, logical inconsistencies, and overconfidence [3].”
Agent Evolution: Capability-Mechanism Co-evolution
Chen, Y., Zhang, C., Zheng, X., & Tang, S. (2026) · International Conference on Generative Artificial Intelligence and Information Security (GAIIS) · doi:10.1109/gaiis69281.2026.11519194
“These models have 4k to 8k token context windows and lack reinforcement alignment for long-horizon tasks, making them prone to hallucinations in zero-shot complex environments during tool selection and state evaluation [6].”
Code-Collaborative Intelligent Query Generation Technology for Complex Relational Data
Tang, C., Lu, Z., Wang, X., et al. (2026) · International Conference on Generative Artificial Intelligence and Information Security (GAIIS) · doi:10.1109/gaiis69281.2026.11519204
“However, as probabilistic generative models, they still struggle with tasks requiring strict logical consistency and factual precision, often exhibiting hallucinations and unstable reasoning [1].”
Liar, Liar, LLM on Fire: Investigating Deception in Turkish Text Generation
Bölücü, N., Yücel, Z., & Çetintaş, D. (2026) · ACM Transactions on Asian and Low-Resource Language Information Processing · doi:10.1145/3811025
A method for function name recovery in binaries via hierarchical semantic propagation
Kong, D.-B., Geng, Y.-Y., Guo, W., Wei, C.-J., Chen, H., Jing, F., He, L., Kong, X., & Wei, Q. (2026) · Cybersecurity · doi:10.1186/s42400-026-00602-6
“…parse_http_header ) is fundamentally different from calling an anonymous function (e.g., sub_4012A0 ) in terms of comprehension cost; for large language models, this difference similarly determines whether stable and concrete semantic judgments can be formed (Jiang et al. 2025; Ji et al. 2023).”
Banik, S., Ahuja, S., Kumari, S., et al. (2026) · International Conference on Networking and Communications (ICNWC) · doi:10.1109/icnwc68145.2026.11518382
Wei, S. T., Chen, K. Z., & Wang, T.-H. (2026) · International Conference on Educational and Information Technology · doi:10.1109/iceit68991.2026.11521597
“However, challenges such as system instability, dataset bias, and cognitive misalignment remain, particularly impacting disadvantaged learners [5, 9, 13].”
Automatic Grading of Chinese as a Second Language Reading Texts Based on Multi-LLM Weighted Voting
Han, X., & Xu, J. (2026) · International Conference on Educational and Information Technology · doi:10.1109/iceit68991.2026.11521725
“Individual LLMs often exhibit stochastic instability, inherent grading biases, and discriminative hallucination–confidently assigning erroneous levels due to corpus-specific bias or lack of domain expertise [13].”
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.




