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
Items 201–360 of 360
Evaluating Semantic Accuracy of Data-to-Text Generation with Natural Language Inference
Dusek, O., & Kasner, Z. (2020) · International Conference on Natural Language Generation · doi:10.48550/arxiv.2011.10819 · Cited by 89
Semantic Noise Matters for Neural Natural Language Generation
Dušek, O., Howcroft, D. M., & Rieser, V. (2019) · Proceedings of the 12th International Conference on Natural Language Generation · doi:10.18653/v1/w19-8652 · Cited by 85 · cited 1×
“Considering this issue, other works [37, 137] correct paired training samples, specifically the input data, according to the references.”
BackgroundImportance of Search and Evaluation Strategies in Neural Dialogue Modeling
Kulikov, I., Miller, A., Cho, K., & Weston, J. (2019) · Proceedings of the 12th International Conference on Natural Language Generation · doi:10.18653/v1/w19-8609 · Cited by 83 · cited 1×
“The PersonaChat datasets [31, 228] have accelerated research into persona consistency [68, 91, 126, 208, 219, 224, 232].”
BackgroundImproving Faithfulness in Abstractive Summarization with Contrast Candidate Generation and Selection
Chen, S., Zhang, F., Sone, K., & Roth, D. (2021) · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2021.naacl-main.475 · Cited by 83
Learning Knowledge Bases with Parameters for Task-Oriented Dialogue Systems
Madotto, A., Cahyawijaya, S., Winata, G. I., Xu, Y., Liu, Z., Lin, Z., & Fung, P. (2020) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2020.findings-emnlp.215 · Cited by 82
Self-citationNeural Path Hunter: Reducing Hallucination in Dialogue Systems via Path Grounding
Dziri, N., Madotto, A., Zaïane, O., & Bose, A. J. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.168 · Cited by 81 · cited 9×
“Such wrong association results in generation with facts mixed up between two similar entities [41, 168].”
BackgroundMethodInfluentialSelf-citationRAGAs: Automated Evaluation of Retrieval Augmented Generation
Es, S., James, J., Espinosa-Anke, L., & Schockaert, S. (2023) · arXiv · doi:10.48550/arxiv.2309.15217 · Cited by 79
On hallucination and predictive uncertainty in conditional language generation
Xiao, Y., & Wang, W. Y. (2021) · Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume · doi:10.18653/v1/2021.eacl-main.236 · Cited by 79
Controlled Hallucinations: Learning to Generate Faithfully from Noisy Data
Filippova, K. (2020) · Findings of the Association for Computational Linguistics: EMNLP 2020 · doi:10.18653/v1/2020.findings-emnlp.76 · Cited by 79
Active skeleton for non-rigid object detection
Bai, X., Wang, X., Latecki, L. J., Liu, W., & Tu, Z. (2009) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2009.5459188 · Cited by 78
(2021) · doi:10.18653/v1/2021.naacl-main.53 · Cited by 78
GO FIGURE: A Meta Evaluation of Factuality in Summarization
Gabriel, S., Celikyilmaz, A., Jha, R., Choi, Y., & Gao, J. (2021) · Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021 · doi:10.18653/v1/2021.findings-acl.42 · Cited by 75 · cited 2×
“[39] further analyze the FEQA and ind that the efectiveness of QA-based metrics depends on the question.”
BackgroundDo Large Language Models Know What They Don’t Know?
Yin, Z., Sun, Q., Guo, Q., Wu, J., Qiu, X., & Huang, X. (2023) · Findings of the Association for Computational Linguistics: ACL 2023 · doi:10.18653/v1/2023.findings-acl.551 · Cited by 75
Do Large Language Models Know What They Don't Know?
Yin, Z., Sun, Q., Guo, Q., Wu, J., Qiu, X., & Huang, X. (2023) · arXiv · doi:10.48550/arxiv.2305.18153 · Cited by 75
Evaluating Correctness and Faithfulness of Instruction-Following Models for Question Answering
Adlakha, V., BehnamGhader, P., Lu, X. H., Meade, N., & Reddy, S. (2023) · arXiv · doi:10.48550/arxiv.2307.16877 · Cited by 74
Increasing Faithfulness in Knowledge-Grounded Dialogue with Controllable Features
Rashkin, H., Reitter, D., Tomar, G. S., & Das, D. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.58 · Cited by 74 · cited 5×
“Controllable generation techniques like control code which could be provided manually [50, 144, 194] or predicted automatically [194], and controlled re-sampling [144] are leveraged to improve faithfulness.”
BackgroundInfluentialNeural Approaches to Conversational AI
Gao, J., Galley, M., & Li, L. (2018) · Proceedings of ACL 2018, Tutorial Abstracts · doi:10.18653/v1/p18-5002 · Cited by 73 · cited 1×
“A task-oriented dialogue system is often composed of several modules: a natural language understanding module, a dialogue manager, and an NLG module [40].”
BackgroundReducing conversational agents' overconfidence through linguistic calibration
Mielke, S. J., Szlam, A., Dinan, E., & Boureau, Y.-L. (2020) · arXiv · doi:10.48550/arxiv.2012.14983 · Cited by 72 · cited 1×
“Whereas the first type is described using the term "consistency" [100, 186, 208] or "coherence" [11, 39], people recently start to call the second type "hallucination" [122, 152].”
BackgroundImproved Natural Language Generation via Loss Truncation
Kang, D., & Hashimoto, T. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.66 · Cited by 70 · cited 3×
“This is similar to the method proposed by Kang and Hashimoto [61], the latter for full sentences in the summarization task.”
BackgroundMethodInfluentialConstrained Decoding for Neural NLG from Compositional Representations in Task-Oriented Dialogue
Balakrishnan, A., Rao, J., Upasani, K., White, M., & Subba, R. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p19-1080 · Cited by 70 · cited 6×
“[6] propose to adopt tree-structured semantic representations and add constraints on decoding, Li et al.”
BackgroundMethodInfluentialTowards Revealing the Mystery behind Chain of Thought: a Theoretical Perspective
Feng, G., Zhang, B., Gu, Y., Ye, H., He, D., & Wang, L. (2023) · arXiv · doi:10.48550/arxiv.2305.15408 · Cited by 69
Factuality Enhanced Language Models for Open-Ended Text Generation
Lee, N., Ping, W., Xu, P., Patwary, M., Fung, P., Shoeybi, M., & Catanzaro, B. (2022) · arXiv · doi:10.48550/arxiv.2206.04624 · Cited by 68 · cited 3×
“[73] show that deliberately added łrandomnessž from sampling-based decoding increases the unexpected nature of the generation and the higher chance of containing hallucinated content.”
BackgroundInfluentialSelf-citationA Simple Recipe towards Reducing Hallucination in Neural Surface Realisation
Nie, F., Yao, J.-G., Wang, J., Pan, R., & Lin, C.-Y. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p19-1256 · Cited by 68 · cited 7×
“Currently, some works [137, 180, 230] point out that using commonsense knowledge can help to gain better numeral representation.”
BackgroundMethodInfluentialDomain Robustness in Neural Machine Translation
Müller, M., Rios, A., & Sennrich, R. (2019) · arXiv · doi:10.48550/arxiv.1911.03109 · Cited by 68
(2022) · doi:10.52202/068431-2506 · Cited by 68
Self-citationSequence-to-Sequence Generation for Spoken Dialogue via Deep Syntax Trees and Strings
Dušek, O., & Jurcicek, F. (2016) · Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) · doi:10.18653/v1/p16-2008 · Cited by 66
Lightman, H., Kosaraju, V., Burda, Y., Edwards, H., Baker, B., Lee, T., Leike, J., Schulman, J., Sutskever, I., & Cobbe, K. (2023) · arXiv · doi:10.48550/arxiv.2305.20050 · Cited by 65
Focus Attention: Promoting Faithfulness and Diversity in Summarization
Aralikatte, R., Narayan, S., Maynez, J., Rothe, S., & McDonald, R. (2021) · doi:10.18653/v1/2021.acl-long.474 · Cited by 64
Plan-then-Generate: Controlled Data-to-Text Generation via Planning
Su, Y., Vandyke, D., Wang, S., Fang, Y., & Collier, N. (2021) · Findings of the Association for Computational Linguistics: EMNLP 2021 · doi:10.18653/v1/2021.findings-emnlp.76 · Cited by 63 · cited 6×
“Similarly, Plan-then-Generate [146] consists of a content planner and a sequence generator and proposes a structure-aware reinforcement learning (RL) objective that encourages the generated output to adhere to the given content plan.”
BackgroundMethodInfluentialA Token-level Reference-free Hallucination Detection Benchmark for Free-form Text Generation
Liu, T., Zhang, Y., Brockett, C., Mao, Y., Sui, Z., Chen, W., & Dolan, B. (2022) · Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2022.acl-long.464 · Cited by 63
The Factual Inconsistency Problem in Abstractive Text Summarization: A Survey
Huang, Y., Feng, X., Feng, X., & Qin, B. (2021) · arXiv · doi:10.48550/arxiv.2104.14839 · Cited by 62 · cited 4×
“These metrics are based on the idea that only the source knowledge reference should entail the entirety of the information in faithful and hallucination-free generation [37, 41, 44, 72, 75, 86, 89, 123, 190].”
BackgroundMethodInfluentialImitation Learning:A Survey of Learning Methods
Hussein, A., Gaber, M. M., Elyan, E., & Jayne, C. (2017) · ACM Computing Surveys · doi:10.1145/3071073 · Cited by 62
MasakhaNER: Named entity recognition for African languages
(2021) · HAL · doi:10.1162/tacl · Cited by 61
Natural Language Generation for Spoken Dialogue System using RNN Encoder-Decoder Networks
Tran, V.-K., & Nguyen, L.-M. (2017) · Conference on Computational Natural Language Learning · doi:10.18653/v1/k17-1044 · Cited by 61 · cited 1×
“Following previous works such as [38, 203], and [185], Li et al.”
MethodCredibility in Information Retrieval
Ginsca, A., Popescu, A., & Lupu, M. (2015) · Foundations and Trends® in Information Retrieval · doi:10.1561/1500000046 · Cited by 60 · cited 1×
“However, this method leads to another research problem – “how to ensure the trustworthiness of the information we use from the web” [58].”
MethodSticking to the Facts: Confident Decoding for Faithful Data-to-Text Generation
Tian, R., Narayan, S., Sellam, T., & Parikh, A. P. (2019) · arXiv · doi:10.48550/arxiv.1910.08684 · Cited by 60 · cited 5×
“[134] propose a conidence score and a variational Bayes training framework to learn the score from data.”
BackgroundInfluentialWeb-based statistical fact checking of textual documents
Magdy, A., & Wanas, N. (2010) · Proceedings of the 2nd international workshop on Search and mining user-generated contents · doi:10.1145/1871985.1872002 · Cited by 59 · cited 1×
“Some literature attempt to use the whole web as the knowledge source [38, 102].”
BackgroundTowards Faithful Neural Table-to-Text Generation with Content-Matching Constraints
Wang, Z., Wang, X., An, B., Yu, D., & Chen, C. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.101 · Cited by 59
Self-Checker: Plug-and-Play Modules for Fact-Checking with Large Language ModelsPreprint
Li, M., Peng, B., Galley, M., Gao, J.-F., & Zhang, Z. (2023) · NAACL-HLT · doi:10.48550/arxiv.2305.14623 · Cited by 58
TRUE: Re-evaluating Factual Consistency Evaluation
Honovich, O., Aharoni, R., Herzig, J., Taitelbaum, H., Kukliansy, D., Cohen, V., Scialom, T., Szpektor, I., Hassidim, A., & Matias, Y. (2022) · Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2022.naacl-main.287 · Cited by 58
Visual Instruction Tuning with Polite FlamingoPreprint
Chen, D., Liu, J., Dai, W., et al. (2023) · AAAI Conference on Artificial Intelligence · doi:10.48550/arxiv.2307.01003 · Cited by 57
Lexically Constrained Neural Machine Translation with Levenshtein Transformer
Susanto, R. H., Chollampatt, S., & Tan, L. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.325 · Cited by 56 · cited 1×
“[32, 70, 140, 161, 166, 197, 198] use constrained decoding to incorporate specific terminology to machine translation, but the above methods can be repurposed to mitigate hallucinations.”
MethodTRUE: Re-evaluating Factual Consistency Evaluation
Honovich, O., Aharoni, R., Herzig, J., Taitelbaum, H., Kukliansy, D., Cohen, V., Scialom, T., Szpektor, I., Hassidim, A., & Matias, Y. (2022) · Proceedings of the Second DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering · doi:10.18653/v1/2022.dialdoc-1.19 · Cited by 56
Discovering Latent Knowledge in Language Models Without Supervision
Burns, C., Ye, H., Klein, D., & Steinhardt, J. (2022) · arXiv · doi:10.48550/arxiv.2212.03827 · Cited by 55
Data-to-text generation with macro planning
Puduppully, R., & Lapata, M. (2021) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00381 · Cited by 53
RealTime QA: What's the Answer Right Now?
Kasai, J., Sakaguchi, K., Takahashi, Y., Bras, R. L., Asai, A., Yu, X., Radev, D., Smith, N. A., Choi, Y., & Inui, K. (2022) · arXiv · doi:10.48550/arxiv.2207.13332 · Cited by 52
Towards Table-to-Text Generation with Numerical Reasoning
Suadaa, L. H., Kamigaito, H., Funakoshi, K., Okumura, M., & Takamura, H. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.115 · Cited by 52 · cited 1×
“Firstly, numbers contain information about scales and are common and crucial in the Data-toText task [175, 230].”
BackgroundRealTime QA: What's the Answer Right Now?
Kasai, J., Sakaguchi, K., Takahashi, Y., Bras, R. L., Asai, A., Yu, X., Radev, D., Smith, N., Choi, Y., & Inui, K. (2023) · Advances in Neural Information Processing Systems 36 · doi:10.52202/075280-2130 · Cited by 52
(2020) · doi:10.18653/v1/2020.findings-emnlp.190 · Cited by 51
Controlling hallucinations at word level in data-to-text generation
Rebuffel, C., Roberti, M., Soulier, L., Scoutheeten, G., Cancelliere, R., & Gallinari, P. (2022) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-021-00801-4 · Cited by 50 · cited 4×
“, parts of the target sentence are loyal to the source input, while others diverge [110].”
BackgroundInfluentialMeasuring Reliability of Large Language Models through Semantic ConsistencyPreprint
Raj, H., Rosati, D., & Majumdar, S. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2211.05853 · Cited by 50
(2020) · doi:10.18653/v1/2020.fever-1.5 · Cited by 49
Evaluating Semantic Accuracy of Data-to-Text Generation with Natural Language Inference
Dušek, O., & Kasner, Z. (2020) · Proceedings of the 13th International Conference on Natural Language Generation · doi:10.18653/v1/2020.inlg-1.19 · Cited by 48 · cited 4×
“[172] Data2Text Dušek and Kasner [39], Liu et al.”
BackgroundInfluentialDialFact: A Benchmark for Fact-Checking in Dialogue
Gupta, P., Wu, C.-S., Liu, W., & Xiong, C. (2022) · Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2022.acl-long.263 · Cited by 48
Self-contradictory Hallucinations of Large Language Models: Evaluation, Detection and Mitigation
Mündler, N., He, J., Jenko, S., & Vechev, M. (2023) · arXiv · doi:10.48550/arxiv.2305.15852 · Cited by 47
Evaluating Attribution in Dialogue Systems: The BEGIN Benchmark
Dziri, N., Rashkin, H., Linzen, T., & Reitter, D. (2022) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00506 · Cited by 47
Evaluating Attribution in Dialogue Systems: The BEGIN Benchmark
Dziri, N., Rashkin, H., Linzen, T., & Reitter, D. (2021) · arXiv · doi:10.48550/arxiv.2105.00071 · Cited by 47
Improving Factual Consistency Between a Response and Persona Facts
Mesgar, M., Simpson, E., & Gurevych, I. (2021) · Conference of the European Chapter of the Association for Computational Linguistics · doi:10.18653/v1/2021.eacl-main.44 · Cited by 46
Operation-guided Neural Networks for High Fidelity Data-To-Text Generation
Nie, F., Wang, J., Yao, J.-G., Pan, R., & Lin, C.-Y. (2018) · Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d18-1422 · Cited by 46
Density Matching for Bilingual Word Embedding
Zhou, C., Ma, X., Wang, D., & Neubig, G. (2019) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/n19-1161 · Cited by 45
Challenges in Building Intelligent Open-domain Dialog SystemsPreprint
Huang, M., Zhu, X., & Gao, J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.05709 · Cited by 44 · cited 2×
“In contrast, the tolerance is relatively higher in dialogue systems because the desired characteristics are not only faithfulness but also user engagement, especially in open-domain dialogue systems [73].”
BackgroundCONFIT: Toward Faithful Dialogue Summarization with Linguistically-Informed Contrastive Fine-tuning
Tang, X., Nair, A., Wang, B., Wang, B., Desai, J., Wade, A., Li, H., Celikyilmaz, A., Mehdad, Y., & Radev, D. (2022) · Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2022.naacl-main.415 · Cited by 43 · cited 1×
“[179] propose another contrastive fine-tuning strategy, named CONFIT, that can improve the factual consistency and overall quality of summaries.”
BackgroundCode-Switching for Enhancing NMT with Pre-Specified TranslationPreprint
Song, K., Zhang, Y., Yu, H., Luo, W., Wang, K., & Zhang, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.09107 · Cited by 43
Let there be a clock on the beach: Reducing Object Hallucination in Image Captioning
Biten, A. F., Gomez, L., & Karatzas, D. (2022) · 2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) · doi:10.1109/wacv51458.2022.00253 · Cited by 43
Chain-of-Verification Reduces Hallucination in Large Language Models
Dhuliawala, S., Komeili, M., Xu, J., Raileanu, R., Li, X., Celikyilmaz, A., & Weston, J. (2023) · arXiv · doi:10.48550/arxiv.2309.11495 · Cited by 42
Incorporating External Knowledge into Machine Reading for Generative Question Answering
Bi, B., Wu, C., Yan, M., Wang, W., Xia, J., & Li, C. (2019) · Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) · doi:10.18653/v1/d19-1255 · Cited by 40
Language Modeling Is Compression
Delétang, G., Ruoss, A., Duquenne, P.-A., Catt, E., Genewein, T., Mattern, C., Grau-Moya, J., Wenliang, L. K., Aitchison, M., Orseau, L., et al. (2023) · arXiv · doi:10.48550/arxiv.2309.10668 · Cited by 39
Modeling Fluency and Faithfulness for Diverse Neural Machine Translation
Feng, Y., Xie, W., Gu, S., Shao, C., Zhang, W., Yang, Z., & Yu, D. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i01.5334 · Cited by 39
Mishra, A., Patel, D., Vijayakumar, A., Li, X. L., Kapanipathi, P., & Talamadupula, K. (2021) · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2021.naacl-main.104 · Cited by 39 · cited 2×
“[131] find that the low performance of NLI-based metrics is mainly caused by the length of the premises in NLI datasets being shorter than the source documents in abstractive summarization.”
BackgroundOpen-Domain Conversational Agents: Current Progress, Open Problems, and Future Directions
Roller, S., Boureau, Y.-L., Weston, J., Bordes, A., Dinan, E., Fan, A., Gunning, D., Ju, D., Li, M., Poff, S., et al. (2020) · arXiv · doi:10.48550/arxiv.2006.12442 · Cited by 38 · cited 3×
“One of the possible reasons for self-contradiction is that current dialogue systems tend to have a short memory of dialogue history [160].”
BackgroundMethodInfluentialSummaC: Re-Visiting NLI-based Models for Inconsistency Detection in Summarization
Laban, P., Schnabel, T., Bennett, P. N., & Hearst, M. A. (2021) · arXiv · doi:10.48550/arxiv.2111.09525 · Cited by 38 · cited 2×
“These metrics are based on the idea that only the source knowledge reference should entail the entirety of the information in faithful and hallucination-free generation [28, 31, 32, 54, 57, 67, 68, 92].”
BackgroundMethodSelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models
Manakul, P., Liusie, A., & Gales, M. J. F. (2023) · arXiv · doi:10.48550/arxiv.2303.08896 · Cited by 38
A Gold Standard Methodology for Evaluating Accuracy in Data-To-Text Systems
Thomson, C., & Reiter, E. (2020) · Proceedings of the 13th International Conference on Natural Language Generation · doi:10.18653/v1/2020.inlg-1.22 · Cited by 38
Read before Generate! Faithful Long Form Question Answering with Machine Reading
Su, D., Li, X., Zhang, J., Shang, L., Jiang, X., Liu, Q., & Fung, P. (2022) · Findings of the Association for Computational Linguistics: ACL 2022 · doi:10.18653/v1/2022.findings-acl.61 · Cited by 37
Self-citationRHO: Reducing Hallucination in Open-domain Dialogues with Knowledge Grounding
Ji, Z., Liu, Z., Lee, N., Yu, T., Wilie, B., Zeng, M., & Fung, P. (2023) · Findings of the Association for Computational Linguistics: ACL 2023 · doi:10.18653/v1/2023.findings-acl.275 · Cited by 37
Self-citationRead before Generate! Faithful Long Form Question Answering with Machine Reading
Su, D., Li, X., Zhang, J., Shang, L., Jiang, X., Liu, Q., & Fung, P. (2022) · arXiv · doi:10.48550/arxiv.2203.00343 · Cited by 37
Self-citation(2019) · doi:10.18653/v1/d19-1292 · Cited by 36 · cited 1×
“However, it has been shown that veriication models are prone to adversarial attacks and are not robust to negation, numerical or comparative words [133].”
BackgroundLanguage Models as Few-Shot Learner for Task-Oriented Dialogue Systems
Madotto, A., Liu, Z., Lin, Z., & Fung, P. (2020) · arXiv · doi:10.48550/arxiv.2008.06239 · Cited by 36 · cited 1×
“Pre-training of models on a large corpus is known to result in the model memorizing knowledge in its parameters [121, 142, 158].”
BackgroundSelf-citationImage inpainting through neural networks hallucinations
Fawzi, A., Samulowitz, H., Turaga, D. S., & Frossard, P. (2016) · Multidimensional Signal Processing Workshop · doi:10.1109/ivmspw.2016.7528221 · Cited by 36
Do Language Embeddings capture Scales?
Zhang, X., Ramachandran, D., Tenney, I., Elazar, Y., & Roth, D. (2020) · Findings of the Association for Computational Linguistics: EMNLP 2020 · doi:10.18653/v1/2020.findings-emnlp.439 · Cited by 35
Differentiable Scheduled Sampling for Credit Assignment
Goyal, K., Dyer, C., & Berg-Kirkpatrick, T. (2017) · Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) · doi:10.18653/v1/p17-2058 · Cited by 34 · cited 1×
“[47] create a diferentiable approximation to greedy decoding that shows a good performance in the NMT task.”
BackgroundLexically Constrained Decoding for Sequence Generation Using Grid Beam SearchPreprint
Hokamp, C., & Liu, Q. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1704.07138 · Cited by 34 · cited 4×
“Liu et al. [94] select training instances based on faithfulness ranking.”
BackgroundInfluentialRevisiting Challenges in Data-to-Text Generation with Fact Grounding
Wang, H. (2019) · Proceedings of the 12th International Conference on Natural Language Generation · doi:10.18653/v1/w19-8639 · Cited by 34 · cited 5×
“” is not mentioned in the source table [194].”
BackgroundInfluentialJoint Parsing and Generation for Abstractive Summarization
Song, K., Lebanoff, L., Guo, Q., Qiu, X., Xue, X., Li, C., Yu, D., & Liu, F. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i05.6419 · Cited by 32
GooAQ: Open Question Answering with Diverse Answer Types
Khashabi, D., Ng, A., Khot, T., Sabharwal, A., Hajishirzi, H., & Callison-Burch, C. (2021) · Findings of the Association for Computational Linguistics: EMNLP 2021 · doi:10.18653/v1/2021.findings-emnlp.38 · Cited by 31 · cited 1×
“It is an important task since many of the everyday questions that humans deal with and pose to search engines require in-depth explanations [81] (e.”
Background(2019) · doi:10.1609/aaai.v33i01.33016618 · Cited by 31
Measuring and Modifying Factual Knowledge in Large Language Models
Pezeshkpour, P. (2023) · arXiv · doi:10.48550/arxiv.2306.06264 · Cited by 30
De-Confounded Variational Encoder-Decoder for Logical Table-to-Text Generation
Chen, W., Tian, J., Li, Y., He, H., & Jin, Y. (2021) · doi:10.18653/v1/2021.acl-long.430 · Cited by 30
Chern, I.-C., Chern, S., Chen, S., Yuan, W., Feng, K., Zhou, C., He, J., Neubig, G., & Liu, P. (2023) · arXiv · doi:10.48550/arxiv.2307.13528 · Cited by 29
KoLA: Carefully Benchmarking World Knowledge of Large Language Models
Yu, J., Wang, X., Tu, S., Cao, S., Zhang-Li, D., Lv, X., Peng, H., Yao, Z., Zhang, X., Li, H., et al. (2023) · arXiv · doi:10.48550/arxiv.2306.09296 · Cited by 29
DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Models
Chuang, Y.-S., Xie, Y., Luo, H., Kim, Y., Glass, J., & He, P. (2023) · arXiv · doi:10.48550/arxiv.2309.03883 · Cited by 29
Certifying LLM Safety against Adversarial Prompting
Kumar, A., Agarwal, C., Srinivas, S., Li, A. J., Feizi, S., & Lakkaraju, H. (2023) · arXiv · doi:10.48550/arxiv.2309.02705 · Cited by 28
Varshney, N., Yao, W., Zhang, H., Chen, J., & Yu, D. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2307.03987 · Cited by 28
Mind The Facts: Knowledge-Boosted Coherent Abstractive Text Summarization
Gunel, B., Zhu, C., Zeng, M., & Huang, X. (2020) · arXiv · doi:10.48550/arxiv.2006.15435 · Cited by 28 · cited 2×
“[65] try to use external knowledge from Wikipedia to make knowledge embeddings.”
BackgroundMethodCognitive Neuroscience: Acting on Numbers
Göbel, S. M., & Rushworth, M. F. S. (2004) · Current Biology · doi:10.1016/j.cub.2004.06.042 · Cited by 27 · cited 1×
“This indiscriminate treatment contradicts findings in cognitive neuroscience, where numbers are known to be represented differently from lexical words in a different part of the brain [60].”
BackgroundHallucination of Speech Recognition Errors With Sequence to Sequence Learning
Serai, P., Sunder, V., & Fosler‐Lussier, E. (2022) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2022.3145313 · Cited by 27 · cited 2×
“We also briely summarize hallucinations in multi-modal settings such as visual-language tasks [1, 10] and speech-to-text tasks [119, 124].”
BackgroundIs Incoherence Surprising? Targeted Evaluation of Coherence Prediction from Language Models
Beyer, A., Loáiciga, S., & Schlangen, D. (2021) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/2021.naacl-main.328 · Cited by 27 · cited 2×
“Two possible types of inconsistency occur in open-domain dialogue generation: (1) self-inconsistency [147, 159] or incoherence [8, 29] among the system utterances, such as when the system contradicts its previous utterance; (2) external inconsistency with an external source, such as factually incorrect utterances.”
BackgroundNeural Machine Translation With Explicit Phrase Alignment
Zhang, J., Luan, H., Sun, M., Zhai, F., Xu, J., & Liu, Y. (2021) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2021.3057831 · Cited by 27
(2019) · doi:10.1609/aaai.v33i01.33016908 · Cited by 26
Improving Large-Scale Fact-Checking using Decomposable Attention Models and Lexical Tagging
Lee, N., Wu, C.-S., & Fung, P. (2018) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/d18-1143 · Cited by 25
Self-citationInvestigating the Catastrophic Forgetting in Multimodal Large Language Models
Zhai, Y., Tong, S., Li, X., Cai, M., Qu, Q., Lee, Y. J., & Ma, Y. (2023) · arXiv · doi:10.48550/arxiv.2309.10313 · Cited by 24
Data-QuestEval: A Referenceless Metric for Data-to-Text Semantic Evaluation
Rebuffel, C., Scialom, T., Soulier, L., Piwowarski, B., Lamprier, S., Staiano, J., Scoutheeten, G., & Gallinari, P. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.633 · Cited by 24 · cited 2×
“Data-QuestEval [111] adapt QuestEval [117] from summarization into data-to-text generation.”
BackgroundMethodRoR: Read-over-Read for Long Document Machine Reading Comprehension
Zhao, J., Bao, J., Wang, Y., Zhou, Y., Wu, Y., He, X., & Zhou, B. (2021) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2021.findings-emnlp.160 · Cited by 22
Do Language Embeddings capture Scales?
Zhang, X., Ramachandran, D., Tenney, I., Elazar, Y., & Roth, D. (2020) · Proceedings of the Third BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP · doi:10.18653/v1/2020.blackboxnlp-1.27 · Cited by 22
D-Score: Holistic Dialogue Evaluation Without Reference
Zhang, C., Lee, G., D'Haro, L. F., & Li, H. (2021) · IEEE/ACM Transactions on Audio, Speech, and Language Processing · doi:10.1109/taslp.2021.3074012 · Cited by 21 · cited 2×
“Recently, several works have proposed evaluation metrics for measuring consistency, such as using natural language inference (NLI) [40, 202], training learnable evaluation metrics [225], or releasing an additional test set for coherence [10].”
BackgroundSanthanam, S., Hedayatnia, B., Gella, S., Padmakumar, A., Kim, S., Liu, Y., & Hakkani-Tur, D. (2021) · arXiv · doi:10.48550/arxiv.2110.05456 · Cited by 21 · cited 6×
“Therefore, chatbots are usually evaluated by humans on factual consistency or factual correctness [154, 198].”
BackgroundMethodInfluentialInvestigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentation
Ren, R., Wang, Y., Qu, Y., Zhao, W. X., Liu, J., Tian, H., Wu, H., Wen, J.-R., & Wang, H. (2023) · arXiv · doi:10.48550/arxiv.2307.11019 · Cited by 21
Aiming to Know You Better Perhaps Makes Me a More Engaging Dialogue Partner
Zemlyanskiy, Y., & Sha, F. (2018) · Conference on Computational Natural Language Learning · doi:10.18653/v1/k18-1053 · Cited by 21 · cited 1×
“PersonaChat datasets [30, 211] accelerate the research on persona consistency [68, 88, 118, 192, 202, 207, 214].”
BackgroundTowards Faithfulness in Open Domain Table-to-text Generation from an Entity-centric View
Liu, T., Zheng, X., Chang, B., & Sui, Z. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i15.17583 · Cited by 21
Towards Enhancing Faithfulness for Neural Machine Translation
Weng, R., Yu, H., Wei, X., & Luo, W. (2020) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.212 · Cited by 21
CIDER: Commonsense Inference for Dialogue Explanation and Reasoning
Ghosal, D., Hong, P., Shen, S., Majumder, N., Mihalcea, R., & Poria, S. (2021) · Proceedings of the 22nd Annual Meeting of the Special Interest Group on Discourse and Dialogue · doi:10.18653/v1/2021.sigdial-1.33 · Cited by 19 · cited 1×
“There are some reasoning works in the area of dialogue [26, 57, 198], but few in reducing hallucinations.”
BackgroundSlot-consistent NLG for Task-oriented Dialogue Systems with Iterative Rectification Network
Li, Y., Yao, K., Qin, L., Che, W., Li, X., & Liu, T. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.10 · Cited by 19 · cited 7×
“Moreover, there are some broad metrics, such as Correctness [6, 11, 99, 186], Accuracy [98, 208], and Informativeness [103] considering both missing and additional contents (extrinsic hallucinations) compared to the input source.”
BackgroundMethodInfluentialCredibility in Information Retrieval
Ginsca, A., Popescu, A., & Lupu, M. (2015) · now publishers, Inc. eBooks · doi:10.1561/9781680830873 · Cited by 18
Briakou, E., & Carpuat, M. J. (2021) · doi:10.18653/v1/2021.acl-long.562 · Cited by 18
Retrieval-Free Knowledge-Grounded Dialogue Response Generation with Adapters
Xu, Y., Ishii, E., Cahyawijaya, S., Liu, Z., Winata, G. I., Madotto, A., Su, D., & Fung, P. (2021) · arXiv · doi:10.48550/arxiv.2105.06232 · Cited by 17 · cited 1×
“Recently, there have been some attempts to generate informative responses without explicit knowledge inputs, but with the help of the implicit knowledge inside large pre-trained language models instead [217, 239] during the inference time.”
BackgroundSelf-citationA Statistical, Grammar-Based Approach to Microplanning
Gardent, C., & Perez-Beltrachini, L. (2016) · Computational Linguistics · doi:10.1162/coli_a_00273 · Cited by 17
Hallucination In Object Detection — A Study In Visual Part VERIFICATION
Kayhan, O. S., Vredebregt, B., & van Gemert, J. C. (2021) · Proceedings - International Conference on Image Processing · doi:10.1109/icip42928.2021.9506670 · Cited by 16
Plan-then-Generate: Controlled Data-to-Text Generation via PlanningPreprint
Su, Y., Vandyke, D., Wang, S., Fang, Y., & Collier, N. (2021) · arXiv (Cornell University) · doi:10.17863/cam.84984 · Cited by 15
Knowledge of Knowledge: Exploring Known-Unknowns Uncertainty with Large Language Models
Amayuelas, A., Wong, K., Pan, L., Chen, W., & Wang, W. (2023) · arXiv · doi:10.48550/arxiv.2305.13712 · Cited by 15
FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation
Min, S., Krishna, K., Lyu, X., Lewis, M., Yih, W.-T., Koh, P. W., Iyyer, M., Zettlemoyer, L., & Hajishirzi, H. (2023) · arXiv · doi:10.48550/arxiv.2305.14251 · Cited by 14
Improving Language Models via Plug-and-Play Retrieval Feedback
Yu, W., Zhang, Z., Liang, Z., Jiang, M., & Sabharwal, A. (2023) · arXiv · doi:10.48550/arxiv.2305.14002 · Cited by 14
(2021) · arXiv (Cornell University) · Cited by 13
Sun, K., Xu, Y. E., Zha, H., Liu, Y., & Dong, X. L. (2023) · arXiv · doi:10.48550/arxiv.2308.10168 · Cited by 13
Self-citationDifferentiable Sampling with Flexible Reference Word Order for Neural Machine Translation
Xu, W., Niu, X., & Carpuat, M. J. (2019) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/n19-1207 · Cited by 13 · cited 1×
“[215] propose further improvement of the scheduled sampling algorithm for NMT by optimizing the probability of source and target word alignments.”
Method(2023) · doi:10.48550/arxiv.2305.13669 · Cited by 13
A Token-level Reference-free Hallucination Detection Benchmark for Free-form Text Generation
Liu, T., Zhang, Y., Brockett, C., Mao, Y., Sui, Z., Chen, W., & Dolan, B. (2021) · arXiv · doi:10.48550/arxiv.2104.08704 · Cited by 12
(2022) · doi:10.48550/arxiv.2210.08726 · Cited by 12
Deduplicating Training Data Makes Language Models Better
Lee, K., Ippolito, D., Nystrom, A., Zhang, C., Eck, D., Callison-Burch, C., & Carlini, N. (2021) · arXiv · doi:10.48550/arxiv.2107.06499 · Cited by 11 · cited 1×
“[96] show that duplicated examples from the pretraining corpus bias the model to favor generating repeats of the memorized phrases from the duplicated examples.”
BackgroundWang, Z., Mao, S., Wu, W., Ge, T., Wei, F., & Ji, H. (2023) · arXiv · doi:10.48550/arxiv.2307.05300 · Cited by 11
Chain of Natural Language Inference for Reducing Large Language Model Ungrounded Hallucinations
Lei, D., Li, Y., Hu, M., Wang, M., Yun, V., Ching, E., & Kamal, E. (2023) · arXiv · doi:10.48550/arxiv.2310.03951 · Cited by 10
Mitigating the Alignment Tax of RLHF
Lin, Y., Lin, H., Xiong, W., Diao, S., Liu, J., Zhang, J., Pan, R., Wang, H., Hu, W., Zhang, H., et al. (2023) · arXiv · doi:10.48550/arxiv.2309.06256 · Cited by 10
(2021) · doi:10.18653/v1/2021.findings-acl.427 · Cited by 10
Sometimes We Want Ungrammatical Translations
Parthasarathi, P., Sinha, K., Pineau, J., & Williams, A. (2021) · Findings of the Association for Computational Linguistics: EMNLP 2021 · doi:10.18653/v1/2021.findings-emnlp.275 · Cited by 9 · cited 1×
“[101] calculate faithfulness by computing similarity scores between perturbed source sentence and target sentence after applying the same perturbation.”
MethodGenerating benchmarks for factuality evaluation of language models
Muhlgay, D., Ram, O., Magar, I., Levine, Y., Ratner, N., Belinkov, Y., Abend, O., Leyton-Brown, K., Shashua, A., & Shoham, Y. (2023) · arXiv · doi:10.48550/arxiv.2307.06908 · Cited by 8
Shen, L., Zhan, H., Shen, X., Chen, H., Zhao, X., & Zhu, X. (2021) · arXiv (Cornell University) · doi:10.1145/3459637.3482352 · Cited by 8
Rule-based Morphological Inflection Improves Neural Terminology Translation
Xu, W., & Carpuat, M. J. (2021) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2021.emnlp-main.477 · Cited by 7
Pretraining Data Mixtures Enable Narrow Model Selection Capabilities in Transformer Models
Yadlowsky, S., Doshi, L., & Tripuraneni, N. (2023) · arXiv · doi:10.48550/arxiv.2311.00871 · Cited by 7
PURR: Efficiently Editing Language Model Hallucinations by Denoising Language Model Corruptions
Chen, A., Pasupat, P., Singh, S., Lee, H., & Guu, K. (2023) · arXiv · doi:10.48550/arxiv.2305.14908 · Cited by 7
Improved Natural Language Generation via Loss TruncationPreprint
Kang, D. D., & Hashimoto, T. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.14589 · Cited by 7
Trusting Your Evidence: Hallucinate Less with Context-aware Decoding
Shi, W., Han, X., Lewis, M., Tsvetkov, Y., Zettlemoyer, L., & Yih, S. W.-T. (2023) · arXiv · doi:10.48550/arxiv.2305.14739 · Cited by 6
Mining the Correlation between Human and Automatic Evaluation at Sentence Level
Sun, Y. (2010) · Proceedings of the Language Resources and Evaluation Conference · doi:10.63317/2cpgtwtr2gg4 · Cited by 6 · cited 1×
“There are two main forms of human evaluation: (1) scoring, where human annotators rate the hallucination level in a range; and (2) comparing, where human annotators compare the output texts with baselines or ground-truth references [176].”
BackgroundQuantifying and Attributing the Hallucination of Large Language Models via Association Analysis
Du, L., Wang, Y., Xing, X., Ya, Y., Li, X., Jiang, X., & Fang, X. (2023) · arXiv · doi:10.48550/arxiv.2309.05217 · Cited by 4
Retrieving Supporting Evidence for LLMs Generated Answers
Huo, S., Arabzadeh, N., & Clarke, C. L. A. (2023) · arXiv · doi:10.48550/arxiv.2306.13781 · Cited by 4
FCM: A Fine-grained Comparison Model for Multi-turn Dialogue Reasoning
Wang, X., Zhang, H., Zhao, S., Zou, Y., Chen, H., Ding, Z., Cheng, B., & Lan, Y. (2021) · Findings of the Association for Computational Linguistics: EMNLP 2021 · doi:10.18653/v1/2021.findings-emnlp.362 · Cited by 4
ExpertQA: Expert-Curated Questions and Attributed Answers
Malaviya, C., Lee, S., Chen, S., Sieber, E., Yatskar, M., & Roth, D. (2023) · arXiv · doi:10.48550/arxiv.2309.07852 · Cited by 2
Instruction Mining: Instruction Data Selection for Tuning Large Language Models
Cao, Y., Kang, Y., Wang, C., & Sun, L. (2023) · arXiv · doi:10.48550/arxiv.2307.06290 · Cited by 2
VScript: Controllable Script Generation with Visual Presentation
Ji, Z., Xu, Y., Cheng, I.-T., Cahyawijaya, S., Frieske, R., Ishii, E., Zeng, M., Madotto, A., & Fung, P. (2022) · arXiv · doi:10.48550/arxiv.2203.00314 · Cited by 2
Self-citationFish, W. (2009) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780195381344.003.0004 · Cited by 2
DialFact: A Benchmark for Fact-Checking in DialoguePreprint
Gupta, P., Wu, C.-S., Liu, W., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.08222 · Cited by 1 · cited 2×
“[66] propose datasets that can benefit fact-checking systems specialized for dialogue systems.”
BackgroundElastic Weight Removal for Faithful and Abstractive Dialogue Generation
Daheim, N., Dziri, N., Sachan, M., Gurevych, I., & Ponti, E. M. (2023) · arXiv · doi:10.48550/arxiv.2303.17574
Think Before You Speak: Explicitly Generating Implicit Commonsense Knowledge for Response Generation
Zhou, P., Gopalakrishnan, K., Hedayatnia, B., Kim, S., Pujara, J., Ren, X., Liu, Y., & Hakkani-Tur, D. (2021) · arXiv · doi:10.48550/arxiv.2110.08501
OpenAI, Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F. L., Almeida, D., Altenschmidt, J., Altman, S., et al. (2023) · arXiv · doi:10.48550/arxiv.2303.08774
Briakou, E., & Carpuat, M. (2021) · arXiv · doi:10.48550/arxiv.2105.15087
Data Augmentation Approaches in Natural Language Processing: A Survey
Li, B., Hou, Y., & Che, W. (2021) · arXiv · doi:10.48550/arxiv.2110.01852
CONFIT: Toward Faithful Dialogue Summarization with Linguistically-Informed Contrastive Fine-tuning
Tang, X., Nair, A., Wang, B., et al. (2021) · arXiv · doi:10.48550/arxiv.2112.08713
A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges
Abdar, M., Pourpanah, F., Hussain, S., et al. (2020) · arXiv · doi:10.48550/arxiv.2011.06225
Rethinking Search: Making Domain Experts out of Dilettantes
Metzler, D., Tay, Y., Bahri, D., et al. (2021) · arXiv · doi:10.48550/arxiv.2105.02274
Survey on reinforcement learning for language processing
Uc-Cetina, V., Navarro-Guerrero, N., Martin-Gonzalez, A., et al. (2021) · arXiv · doi:10.48550/arxiv.2104.05565
CIDER: Commonsense Inference for Dialogue Explanation and Reasoning
Ghosal, D., Hong, P., Shen, S., et al. (2021) · arXiv · doi:10.48550/arxiv.2106.00510
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
Accurate structure prediction of biomolecular interactions with AlphaFold 3
Abramson, J., Adler, J., Dunger, J., Evans, R., Green, T., Pritzel, A., Ronneberger, O., Willmore, L., Ballard, A. J., Bambrick, J., et al. (2024) · Nature · doi:10.1038/s41586-024-07487-w · Cited by 15,861
“The biggest issue is that generative models are prone to hallucination 35 , whereby the model may invent plausible-looking structure even in unstructured regions.”
BackgroundLarge language models in medicine
Thirunavukarasu, A. J., Ting, D. S. J., Elangovan, K., Gutierrez, L., Tan, T. F., & Ting, D. S. W. (2023) · Nature Medicine · doi:10.1038/s41591-023-02448-8 · Cited by 3,944
“‘Watermarking’ protocols could facilitate high-quality text generation with detectable signatures signaling LLM involvement, but this is not currently being implemented in the most popular models 109 .”
BackgroundHuang, L., Yu, W., Ma, W., Zhong, W., Feng, Z., Wang, H., Chen, Q., Peng, W., Feng, X., Qin, B., et al. (2025) · ACM Transactions on Information Systems · doi:10.1145/3703155 · Cited by 4,688
“The current definition of hallucinations aligns with prior research (Ji et al., 2023a), characterizing them as generated content that is nonsensical or unfaithful to the provided source content.”
InfluentialA Survey on Large Language Model based Autonomous Agents
Wang, L., Ma, C., Feng, X., Zhang, Z., Yang, H., Zhang, J., Chen, Z., Tang, J., Chen, X., Lin, Y., et al. (2023) · arXiv · doi:10.48550/arxiv.2308.11432 · Cited by 1,925
A survey on large language model based autonomous agents
Wang, L., Ma, C., Feng, X., Zhang, Z., Yang, H., Zhang, J., Chen, Z., Tang, J., Chen, X., Lin, Y., et al. (2024) · Frontiers of Computer Science · doi:10.1007/s11704-024-40231-1 · Cited by 1,941
“Hallucination can lead to serious consequences such as incorrect or misleading code, security risks, and ethical issues [74].”
Background
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.




