Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty
Open accessThe DOI matches its Crossref record; 223 of the 229 references it cites were found in a registry or an index, and 6 are not checked yet; 41 of the 135 sources its claims cite were checked for support.
Abstract
Algorithmic transparency entails exposing system properties to various stakeholders for purposes that include understanding, improving, and contesting predictions. Until now, most research into algorithmic transparency has predominantly focused on explainability. Explainability attempts to provide reasons for a machine learning model's behavior to stakeholders. However, understanding a model's specific behavior alone might not be enough for stakeholders to gauge whether the model is wrong or lacks sufficient knowledge to solve the task at hand. In this paper, we argue for considering a complementary form of transparency by estimating and communicating the uncertainty associated with model predictions. First, we discuss methods for assessing uncertainty. Then, we characterize how uncertainty can be used to mitigate model unfairness, augment decision-making, and build trustworthy systems. Finally, we outline methods for displaying uncertainty to stakeholders and recommend how to collect information required for incorporating uncertainty into existing ML pipelines. This work constitutes an interdisciplinary review drawn from literature spanning machine learning, visualization/HCI, design, decision-making, and fairness. We aim to encourage researchers and practitioners to measure, communicate, and use uncertainty as a form of transparency.
Verified 2026-11-19 · 6 references not checked yet
Overview
CoChat Summary
TL;DR
This paper argues that estimating and communicating uncertainty should be treated as a complementary form of algorithmic transparency alongside explainability. It reviews methods for measuring uncertainty and discusses how uncertainty can support fairness, decision-making, trust, and stakeholder communication.
Key findings
- Uncertainty is split into aleatoric uncertainty (irreducible noise/class overlap) and epistemic uncertainty (reducible with more data); calibration is a precursor to using uncertainty as transparency, and deep ensembles are described as one of the best-performing uncertainty quantification approaches for neural networks, retaining calibration under dataset shift.
- Uncertainty can help address fairness issues: representation bias is epistemic and may be detected via large epistemic uncertainty on validation/deployment data, while feature and label noise are aleatoric and can distort fairness assessments; even a small amount of label noise can greatly impact fairness evaluation.
- In decision-making, uncertainty supports reject options and decision-theoretic action selection; JDM findings indicate risk is considered together with expected outcome, asymmetrically, and tolerance for uncertainty decreases superlinearly as stakes increase.
- Communication is difficult: in a 2010 US/Germany survey, 20% could not identify 10% as the biggest disease risk among 1%, 5%, and 10%, and almost 30% could not answer whether 1 in 10, 1 in 100, or 1 in 1000 represents the largest risk; people interpreted the IPCC’s “very likely” as around 60% probability, whereas the guideline specifies greater than 90%. Quantile dot plots and hypothetical outcome plots (HOPs) are reported to improve some probability/uncertainty judgments.
Methods
- Interdisciplinary literature review, not an empirical study, drawn from machine learning, visualization/HCI, design, decision-making, and fairness.
- Uses a recurring breast cancer diagnostics scenario to illustrate uncertainty in fairness, decision-making, trust formation, and communication.
- Synthesizes uncertainty quantification approaches (Bayesian methods, frequentist methods such as ensembles), evaluation metrics (calibration, NLL, Brier score, ECE), and communication methods (categories, numerical ranges, summary statistics, full distributions, visualizations).
Limitations
- The paper states that its findings should not be treated as conclusive for predicting real-world usability and recommends testing with real users in concrete tasks and settings.
- The paper states that empirical understanding of how decision-makers use aleatoric versus epistemic uncertainty is limited, that JDM literature has mostly focused on discrete outcomes, and that there is not a good understanding of how people perceive uncertainty over continuous outcomes such as error bars.
- Evident from the text: as a review, it provides no primary empirical validation or systematic review protocol; several claims are speculative or drawn from other fields and are framed as open questions for future research.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures




![(c) Cone of uncertainty, showing the uncertainty around the predicted path of the center of a hurricane. Taken from [93].](https://figures.cochat.ai/paper-figures/42e1c0dd14b540788474c4bee4ed993b/uncertainty-as-a-form-of-transparency-measuring-communicating-and-using-figure-5.png)

Verification
Claims the paper makes
28 Valid · 10 Not valid · 60 Not checked · 104 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“When information on the sensitive attribute is unavailable (e.g., information on gender was not collected), but can be predicted from an auxiliary dataset, disparity measures are generally unidentifiable [20,63].”
§ Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty > 3.1 Uncertainty and Fairness
What the sources say
- Partly supportedcites [20] 20. Jiahao Chen, Nathan Kallus, Xiaojie Mao, Geoffry Svacha, and Madeleine Udell. 2019. Fairness under unawareness: Assessing disparity when protected class is unobserved. In Proceedings of the conference on fairness, accountability, and transparency. 339-348. sourceExperimental — accuracy not yet measured
Section 1.1 (p. 2) describes assessing disparity when protected-class labels are unavailable and must be imputed from proxy variables. The paper analyzes bias in the resulting disparity estimates, but the cited passages do not establish the broader claim that disparity measures are generally unidentifiable; the Conclusions (p. 9) instead note that the optimal imputation threshold cannot be determined without ground-truth labels. (LLM, conf=0.82)
- Supportedcites [63] 63. Nathan Kallus, Xiaojie Mao, and Angela Zhou. 2019. Assessing algorithmic fairness with unobserved protected class using data combination. arXiv preprint arXiv:1906.00285 (2019). source
The paper’s problem formulation describes a primary dataset missing protected-class labels and an auxiliary dataset containing those labels and proxy variables. Section 4 states that disparity measures are generally unidentifiable from the two datasets because their marginals do not uniquely determine the full joint distribution. The paper notes an exception for perfectly predictive proxies, consistent with the claim’s qualifier ‘generally.’ (LLM, conf=0.94)
plain- 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.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The experimental results of Gupta et al. [44] have shown that enforcing fairness constraints on a noisy sensitive attribute, without assumptions on the structure of the noise, is not guaranteed to lead to any improvement in the fairness properties of our model.”
§ Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty > 3.1 Uncertainty and Fairness
What the sources say
- Partly supportedcites [44] 44. Maya Gupta, Andrew Cotter, Mahdi Milani Fard, and Serena Wang. 2018. Proxy fairness. arXiv preprint arXiv:1806.11212 (2018). sourceExperimental — accuracy not yet measured
Section 4 (pp. 3–4) says poorly chosen proxy groups can worsen fairness for the true protected groups, supporting the claim that improvement is not guaranteed. But the paper presents this as a theoretical possibility, while its reported experiments generally found improvements (Conclusions, p. 8). It does not frame the result as an experiment on noisy sensitive attributes without assumptions about the noise structure. (LLM, conf=0.84)
boosted- 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.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Indeed, Fogliato et al. [35] show that even a small amount of label noise can greatly impact the assessment of the fairness properties of the model.”
§ Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty > 3.1 Uncertainty and Fairness
What the sources say
- Supportedcites [35] 35. Riccardo Fogliato, Max G'Sell, and Alexandra Chouldechova. 2020. Fairness Evaluation in Presence of Biased Noisy Labels. arXiv preprint arXiv:2003.13808 (2020). source
The Abstract (p. 1) reports that even small biases in observed labels may call into question conclusions drawn from evaluating a model against those labels. The Conclusion (p. 8) gives an example in which a small gap in arrest likelihood could fully account for observed miscalibration. This supports the claim that a small amount of biased label noise can substantially affect fairness assessment. (LLM, conf=0.89)
plainstates a 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
Every source the paper cites for this claim was checked and supports it.
In the paper
“Here, measures of demographic parity and equalized odds computed on the observed data are equal to the true metrics up to a scaling factor, which is proportional to the value of the noise [71]; if the noise rates are known, then the true metrics can be directly estimated.”
§ Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty > 3.1 Uncertainty and Fairness
What the sources say
- Supportedcites [71] 71. Alex Lamy, Ziyuan Zhong, Aditya K Menon, and Nakul Verma. 2019. Noise- tolerant fair classification. In Advances in Neural Information Processing Systems. 294-306. source
Section 4.2 (p. 6) states that MC noise scales the fairness measure, giving the clean measure as (1 − α − β) times the corrupted-data measure. Section 3.3 (p. 4) says equalized odds can be captured by adding a constraint for Y = 0, and §4.3 (p. 6) explains how known noise parameters supply the scaling factor. The factor depends on the noise rates; more precisely, it is 1 − α − β, not directly proportional to their value. (LLM, conf=0.88)
plainstates an association- 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
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Similarly to noise in the sensitive attribute, Blum and Stangl [11], Jiang and Nachum [59] show that fairness constraints are guaranteed to improve the properties of a predictor only under appropriate assumptions on the label noise.”
§ Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty > 3.1 Uncertainty and Fairness
What the sources say
- Partly supportedcites [11] 11. Avrim Blum and Kevin Stangl. 2019. Recovering from biased data: Can fairness constraints improve accuracy? arXiv preprint arXiv:1912.01094 (2019). sourceExperimental — accuracy not yet measured
Section 3.3 (p. 11) says Equal Opportunity-constrained ERM recovers the Bayes-optimal classifier when Group A has sufficient mass and the signal is not too noisy. Section 3.1 (p. 9) also gives a recovery range bounded by the noise parameter η. The claim captures the dependence on noise assumptions, but overgeneralizes the guarantee to fairness constraints generally and omits the paper’s other conditions. (LLM, conf=0.86)
- Partly supportedcites [59] 59. Heinrich Jiang and Ofir Nachum. 2020. Identifying and correcting label bias in machine learning. In International Conference on Artificial Intelligence and Statistics. 702-712. sourceExperimental — accuracy not yet measured
Section 2.1 (p. 2) says that recovering unbiased labels requires an assumption about their relationship to observed biased labels. The Introduction (p. 1) reports a theoretical guarantee for the paper’s re-weighted training method under its model. It does not establish the broader claim that fairness constraints improve a predictor only under appropriate assumptions about label noise. (LLM, conf=0.83)
boostedstates a possible 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
74 more claims · 11 with their cited sources checked
Create a free account to see all 202 claims and to verify the 24 references still open.
- It may state two separate claims as one. 40
- Probably not a substantial claim. 105
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 cited source says what the claim attributes to it.
In the citing paper
“Uncertainty is considered a critical form of transparency that can alert people about the limitations of AI [10].”
From Investigating Explainability of Generative AI for Code through Scenario-based Design (2022) · cites this paper as [10]
The cited source says what the claim attributes to it.
In the citing paper
“Similar metrics have been used to evaluate the calibration of uncertainty quantification [7,9,64]-the estimated uncertainty should reflect the observed error rate.”
From Understanding the Role of Human Intuition on Reliance in Human-AI Decision-Making with Explanations (2023) · cites this paper as [7]
The cited source says what the claim attributes to it.
In the citing paper
“Uncertainty expression can also increase the system's perceived transparency [13], which is generally shown to enhance trust, whether or not appropriate [56].”
From "I'm Not Sure, But...": Examining the Impact of Large Language Models' Uncertainty Expression on User Reliance and Trust (2024) · cites this paper as [13]
The cited source says what the claim attributes to it.
In the citing paper
“Finally, indicating model uncertainty [19,174,198], for example, by means of color-coding [167], and an explanation for that uncertainty [5,190], can also help users disentangle the role of their prompting and output evaluation ability from that of models' capabilities, further supporting confidence adjustment [197].”
From The Metacognitive Demands and Opportunities of Generative AI (2024) · cites this paper as [19]
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
§ Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty > References
“137. Yunfeng Zhang, Rachel KE Bellamy, and Wendy A. Kellogg. 2012.”
§ Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty > References
“In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems (2015).”
Citations
Cites
A Mathematical Theory of Communication
Shannon, C. E. (1948) · Bell System Technical Journal · doi:10.1002/j.1538-7305.1948.tb01338.x · Cited by 88,486
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805
Tversky, A., & Kahneman, D. (1974) · Science · doi:10.1126/science.185.4157.1124 · Cited by 28,810
UCI Machine Learning Repository
Asuncion, A. (2007) · Medical Entomology and Zoology · Cited by 25,370
Judgment under Uncertainty: Heuristics and Biases
Tversky, A., & Kahneman, D. (1975) · Utility, Probability, and Human Decision Making · doi:10.1007/978-94-010-1834-0_8 · Cited by 22,650
"Why Should I Trust You?": Explaining the Predictions of Any Classifier
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939778 · Cited by 17,555
Knight, F. H. (1921) · Cited by 16,535
Breiman, L. (1996) · Machine Learning · doi:10.1007/bf00058655 · Cited by 16,325
Advances in prospect theory: Cumulative representation of uncertainty
Tversky, A., & Kahneman, D. (1992) · Journal of Risk and Uncertainty · doi:10.1007/bf00122574 · Cited by 15,467 · cited 2×
“Research also found that this bias occurs when people are asked to make decisions under risk and that their decisions imply such distortions [125, 137].”
Background(2011) · Cited by 14,317
Research works printed but not linked: 11
- 6. Emily Badger, Claire Cain Miller, Adam Pearce, and Kevin Quealy. 2018. Income Mobility Charts for Girls, Asian-Americans and Other Groups. Or Make Your Own. (2018).
- 19. Shelly Chaiken. 1999. The heuristic-systematic. Dual-process theories social psychology 73 (1999).
- 23. Marvin S Cohen, Raja Parasuraman, and Jared T Freeman. 1998. Trust in decision aids: A model and its training implications. In in Proc. Command and Control Research and Technology Symp. Citeseer.
- 36. Yarin Gal. 2016. Uncertainty in deep learning. University of Cambridge 1, 3 (2016).
- 54. Carl Iver Hovland, Irving Lester Janis, and Harold H Kelley. 1953. Communica- tion and persuasion. (1953).
- 87. Malik Sajjad Ahmed Nadeem, Jean-Daniel Zucker, and Blaise Hanczar. 2009. Accuracy-Rejection Curves (ARCs) for Comparing Classification Methods with a Reject Option. In Proceedings of the third International Workshop on Machine Learning in Systems Biology (Proceedings of Machine Learning Research), Sašo Džeroski, Pierre Guerts, and Juho Rousu (Eds.), Vol. 8. PMLR, Ljubljana, Slovenia, 65-81. http://proceedings.mlr.press/v8/nadeem10a.html
- 91. Radford M Neal. 1995. Bayesian Learning for Neural Networks. PhD thesis, University of Toronto (1995).
- 103. Jennifer Preece, Helen Sharp, and Yvonne Rogers. 2015. Interaction design: beyond human-computer interaction. John Wiley & Sons.
- 118. S Shyam Sundar. 2008. The MAIN model: A heuristic approach to understanding technology effects on credibility. MacArthur Foundation Digital Media and Learning Initiative.
- 127. Anne Marthe van der Bles, Sander van der Linden, Alexandra L. J. Freeman, James Mitchell, Ana B. Galvao, Lisa Zaval, and David J. Spiegelhalter. 2019. Communicating Uncertainty about Facts, Numbers and Science. 6, 5 (2019), 181870. https://doi.org/10/gf2g9j
- 130. Claus O. Wilke. 2019. Fundamentals of Data Visualization: A Primer on Making Informative and Compelling Figures (1st edition ed.). O'Reilly Media.
Cited by
Generative AI and ChatGPT: applications, challenges, and AI-human collaboration
Fui-Hoon Nah, F., Zheng, R., Cai, J., Siau, K., & Chen, L. (2023) · Journal of Information Technology Case and Application Research · doi:10.1080/15228053.2023.2233814 · Cited by 1,363
The Metacognitive Demands and Opportunities of Generative AI
Tankelevitch, L., Kewenig, V., Simkute, A., Scott, A. E., Sarkar, A., Sellen, A., & Rintel, S. (2024) · Proceedings of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642902 · Cited by 493
Towards trustworthy machine fault diagnosis: A probabilistic Bayesian deep learning framework
Zhou, T., Han, T., & Droguett, E. L. (2022) · Reliability Engineering & System Safety · doi:10.1016/j.ress.2022.108525 · Cited by 260
Socially Responsible AI Algorithms: Issues, Purposes, and Challenges
Cheng, L., Varshney, K. R., & Liu, H. (2021) · Journal of Artificial Intelligence Research · doi:10.1613/jair.1.12814 · Cited by 261
Nemani, V., Biggio, L., Huan, X., Hu, Z., Fink, O., Tran, A., Wang, Y., Zhang, X., & Hu, C. (2023) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2023.110796 · Cited by 243
How transparency modulates trust in artificial intelligence
Zerilli, J., Bhatt, U., & Weller, A. (2022) · Patterns · doi:10.1016/j.patter.2022.100455 · Cited by 257
Self-citationData Cards: Purposeful and Transparent Dataset Documentation for Responsible AI
Pushkarna, M., Zaldivar, A., & Kjartansson, O. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3531146.3533231 · Cited by 222
“This builds greater trust in the dataset and subsequently, its publishers [10].”
BackgroundThree Challenges for AI-Assisted Decision-Making
Steyvers, M., & Kumar, A. (2024) · Perspectives on Psychological Science · doi:10.1177/17456916231181102 · Cited by 217
“Most AI systems can calculate confidence measures for their predictions, such as confidence intervals for regression tasks or estimated probabilities of accurate predictions in classification tasks (Bhatt et al., 2021).”
Investigating Explainability of Generative AI for Code through Scenario-based Design
Sun, J., Liao, Q. V., Muller, M., Agarwal, M., Houde, S., Talamadupula, K., & Weisz, J. D. (2022) · 27th International Conference on Intelligent User Interfaces · doi:10.1145/3490099.3511119 · Cited by 219
“We built upon this work, as well as other work that explores XAI or transparency features [10, 21, 31], to select and adapt features that could apply to the context of GenAI for code.”
BackgroundMethodInfluentialSelf-citationUnderstanding the Role of Human Intuition on Reliance in Human-AI Decision-Making with Explanations
Chen, V., Liao, Q. V., Wortman Vaughan, J., & Bansal, G. (2023) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3610219 · Cited by 220
“A similar metric has been used to evaluate the calibration of uncertainty quantification [7]—the reported uncertainty should be highly correlated with the observed error rate.”
MethodSelf-citationREFORMS: Consensus-based Recommendations for Machine-learning-based Science
Kapoor, S., Cantrell, E. M., Peng, K., Pham, T. H., Bail, C. A., Gundersen, O. E., Hofman, J. M., Hullman, J., Lones, M. A., Malik, M. M., et al. (2024) · Science Advances · doi:10.1126/sciadv.adk3452 · Cited by 129
“In addition, communicating and reasoning about uncertainty is important, but uncertainty is currently under-reported ( 169 – 173 ).”
Bridging the Gulf of Envisioning: Cognitive Challenges in Prompt Based Interactions with LLMs
Subramonyam, H., Pea, R., Pondoc, C., Agrawala, M., & Seifert, C. (2024) · Proceedings of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642754 · Cited by 162
“Overall, explainability, or the lack thereof, is a signifcant contributor to the gaps involved in LLM interaction, as users struggle to build a mental model [15, 170, 174].”
BackgroundKim, S. S. Y., Liao, Q. V., Vorvoreanu, M., Ballard, S., & Vaughan, J. W. (2024) · The 2024 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3630106.3658941 · Cited by 147
Self-citationCooper, A. F., Moss, E., Laufer, B., & Nissenbaum, H. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3531146.3533150 · Cited by 129
“property of models, particularly with regard to whether a model is interpretable or explainable to relevant stakeholders [14, 38, 49].”
BackgroundAn uncertainty-informed framework for trustworthy fault diagnosis in safety-critical applications
Zhou, T., Zhang, L., Han, T., Droguett, E. L., Mosleh, A., & Chan, F. T. S. (2022) · Reliability Engineering & System Safety · doi:10.1016/j.ress.2022.108865 · Cited by 118
Prabhudesai, S., Yang, L., Asthana, S., Huan, X., Liao, Q. V., & Banovic, N. (2023) · Proceedings of the 28th International Conference on Intelligent User Interfaces · doi:10.1145/3581641.3584033 · Cited by 82
“Transparency mechanisms [28, 30], including uncertainty communication [13], enable decision-makers to understand model limitations and have appropriate trust and reliance on AI [74].”
BackgroundInfluentialSelf-citationFan, X., Zhang, X., & Yu, X. B. (2023) · Reliability Engineering & System Safety · doi:10.1016/j.ress.2023.109088 · Cited by 82
Look Before You Leap: An Exploratory Study of Uncertainty Analysis for Large Language Models
Huang, Y., Song, J., Wang, Z., Zhao, S., Chen, H., Juefei-Xu, F., & Ma, L. (2025) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2024.3519464 · Cited by 339
“…open source initiatives, NGOs, and businesses may all be negatively affected by untrustworthy LLMs. Uncertainty estimation, aimed at gauging the confidence level of model outputs [26], [27], [28], stands out as a promising approach for identifying risks in general Machine Learning (ML) models.”
Singh, Y., Hathaway, Q. A., Keishing, V., Salehi, S., Wei, Y., Horvat, N., Vera-Garcia, D. V., Choudhary, A., Mula Kh, A., Quaia, E., et al. (2025) · Bioengineering · doi:10.3390/bioengineering12080879 · Cited by 72
“A model predicting pneumonia with 51% probability should be treated fundamentally differently from one predicting the same condition with 95% probability, yet current XAI visualizations often present both scenarios with equal visual confidence [37].”
Lee, J., Herskovitz, J., Peng, Y.-H., & Guo, A. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3501966 · Cited by 61
“As mentioned, one potential method for increasing skepticism in captions is to verbally frame them as uncertain [35], which is a form of model transparency [3].”
MethodLiao, Q. V., Zhang, Y., Luss, R., Doshi-Velez, F., & Dhurandhar, A. (2022) · Proceedings of the AAAI Conference on Human Computation and Crowdsourcing · doi:10.1609/hcomp.v10i1.21995 · Cited by 69
Self-citationExplainability and white box in drug discovery
Kırboğa, K. K., Abbasi, S. W., & Küçüksille, E. U. (2023) · Chemical Biology & Drug Design · doi:10.1111/cbdd.14262 · Cited by 68
“By giving users uncertainty estimates on counterfactual explanations, we may avoid giving them overconfident and perhaps dangerous options, which can help them make better decisions and increase their trust in intelligent systems (Bhatt et al., 2021; Jesson et al., 2020).”
BackgroundSubramanian, H. V., Canfield, C., & Shank, D. B. (2024) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2024.102780 · Cited by 84
Abdar, M., Khosravi, A., Islam, S. M. S., Acharya, U. R., & Vasilakos, A. V. (2022) · IEEE Systems, Man, and Cybernetics Magazine · doi:10.1109/msmc.2022.3150144 · Cited by 55
Ilori, O., Nwosu, N. T., & Naiho, H. N. N. (2024) · Finance & Accounting Research Journal · doi:10.51594/farj.v6i6.1213 · Cited by 76
Fair Conformal Predictors for Applications in Medical Imaging
Lu, C., Lemay, A., Chang, K., Hoebel, K., & Kalpathy–Cramer, J. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i11.21459 · Cited by 150
“To facilitate greater trust in black box models for medical decision-making, uncertainty quantification frameworks have been promoted and used to detect bias in protected patient demographics (Bhatt et al. 2021; Kompa, Snoek, and Beam 2021; Lu et al. 2021b).”
MethodMa, S., Zhou, T., Nie, F., & Ma, X. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3517482 · Cited by 51
“The reasons can be categorized as follows: 1) It was the instructors’ initiative to observe the state of students instead of a “ have to ” behavior, so they “ didn’t think it was a burden ”(T3, 8, 9, 10, 12, 14, 17).”
BackgroundA Critical Survey on Fairness Benefits of Explainable AI
Deck, L., Schoeffer, J., De‐Arteaga, M., & Kühl, N. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3658990 · Cited by 144
Evaluating language models for mathematics through interactions
Collins, K. M., Jiang, A. Q., Frieder, S., Wong, L., Zilka, M., Bhatt, U., Lukasiewicz, T., Wu, Y., Tenenbaum, J. B., Hart, W., et al. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2318124121 · Cited by 49
Self-citationBayesian autoencoders with uncertainty quantification: Towards trustworthy anomaly detection
Yong, B. X., & Brintrup, A. M. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.118196 · Cited by 65
Mehdiyev, N., Majlatow, M., & Fettke, P. (2024) · Annals of Operations Research · doi:10.1007/s10479-024-05943-4 · Cited by 69
“UQ can be perceived as a supplementary form of transparency, which enhances the understandability of resolutions for decision-making tasks, an aspect potentially not addressed sufficiently by other means [61].”
He, G., Buijsman, S., & Gadiraju, U. (2023) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3610067 · Cited by 39
“[4] to use uncertainty measures to help users rely appropriately on AI systems.”
MethodMehdiyev, N., Majlatow, M., & Fettke, P. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.110363 · Cited by 53
A study of UX practitioners roles in designing real-world, enterprise ML systems
Zdanowska, S., & Taylor, A. S. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3517607 · Cited by 46
“A significant strand of work focuses on the low-level interactive aspects of use with AI/ML systems and here considerable focus is given to improving the intelligibility, transparency and explainability of models and algorithms that are often inaccessible to users [2, 9, 21, 52].”
BackgroundPerspectives of Digital Marketing for the Restaurant Industry
Talukder, M. B., Kumar, S., & Das, I. R. (2024) · Advances in media, entertainment and the arts (AMEA) book series · doi:10.4018/979-8-3693-0855-4.ch009 · Cited by 47
Never tell me the odds: Investigating pro-hoc explanations in medical decision making
Cabitza, F., Natali, C., Famiglini, L., Campagner, A., Caccavella, V., & Gallazzi, E. (2024) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2024.102819 · Cited by 35
Uncertainty Quantification with Pre-trained Language Models: A Large-Scale Empirical Analysis
Xiao, Y., Liang, P. P., Bhatt, U., Neiswanger, W., Salakhutdinov, R., & Morency, L.-P. (2022) · Findings of the Association for Computational Linguistics: EMNLP 2022 · doi:10.18653/v1/2022.findings-emnlp.538 · Cited by 32
Self-citationUncertainty Quantification in Geochemical Mapping: A Review and Recommendations
Wang, J., & Zuo, R. (2024) · Geochemistry Geophysics Geosystems · doi:10.1029/2023gc011301 · Cited by 44
Everyday Uncertainty: How Blind People Use GenAI Tools for Information Access
Tang, X., Abdolrahmani, A., Gergle, D., & Piper, A. M. (2025) · Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713433 · Cited by 30
Uncertainty in XAI: Human Perception and Modeling Approaches
Chiaburu, T., Haußer, F., & Bießmann, F. (2024) · Machine Learning and Knowledge Extraction · doi:10.3390/make6020055 · Cited by 53
“Uncertainty, as defined by [68], refers to a “lack of knowledge about some outcome”.”
Background“Role of Human-AI interaction in selective prediction,”
Bondi, E., Koster, R., Sheahan, H., Chadwick, M., Bachrach, Y., Cemgil, T., Paquet, U., & Dvijotham, K. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i5.20465 · Cited by 30
“Bhatt et al. (2021) find that humans may unreliably interpret uncertainty estimates, but that the estimates may increase transparency and thereby performance.”
BackgroundSocially Responsible AI Algorithms: Issues, Purposes, and ChallengesPreprint
Cheng, L., Varshney, K. R., & Liu, H. (2021) · Journal of Artificial Intelligence Research · doi:10.48550/arxiv.2101.02032 · Cited by 249
Bridging the gap: Towards an expanded toolkit for AI-driven decision-making in the public sector
Fischer-Abaigar, U., Kern, C. A., Barda, N., & Kreuter, F. (2024) · Government Information Quarterly · doi:10.1016/j.giq.2024.101976 · Cited by 81
Encoding the Latent Posterior of Bayesian Neural Networks for Uncertainty Quantification
Franchi, G., Bursuc, A., Aldea, E., Dubuisson, S., & Bloch, I. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3328829 · Cited by 50
“of the sources and forms of uncertainty [1], [6], [7], which in turn allows for a better communication of the decisions to humans [8], [9] or to connected modules in an autonomous system [10].”
BackgroundPath To Gain Functional Transparency In Artificial Intelligence With Meaningful Explainability
Hosain, M. T., ANİK, M. H., RAFİ, S., TABASSUM, R., Insia, K., & SIDDIKY, M. M. (2023) · Journal of Metaverse · doi:10.57019/jmv.1306685 · Cited by 48
“Probability is seen by the Bayesian school of thinking as individual levels of conviction that a desired result will materialize [16] [2].”
BackgroundMethodInfluentialEliciting and Learning with Soft Labels from Every Annotator
Collins, K. M., Bhatt, U., & Weller, A. (2022) · Proceedings of the AAAI Conference on Human Computation and Crowdsourcing · doi:10.1609/hcomp.v10i1.21986 · Cited by 74
Self-citationIlori, O., Nwosu, N. T., & Naiho, H. N. N. (2024) · Computer Science & IT Research Journal · doi:10.51594/csitrj.v5i6.1224 · Cited by 41
Yang, T., Qiao, Y., & Lee, B. (2024) · Computers & Security · doi:10.1016/j.cose.2024.103909 · Cited by 44
Lu, C., Angelopoulos, A. N., & Pomerantz, S. R. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-16452-1_52 · Cited by 71
Enhancing IT audit effectiveness with agile methodologies: A conceptual exploration
Ilori, O., Nwosu, N. T., & Naiho, H. N. N. (2024) · Engineering Science & Technology Journal · doi:10.51594/estj.v5i6.1217 · Cited by 32
Robot-Assisted Decision-Making: Unveiling the Role of Uncertainty Visualisation and Embodiment
Schömbs, S., Pareek, S., Gonçalves, J., & Johal, W. (2024) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642911 · Cited by 41
“Similarly, Bhatt et al. [9] stresses the importance of communicating uncertainty to enhance transparency and to support an appropriate formation of trust between the agent and the human.”
BackgroundResultInfluentialNavigating Uncertainty: A User-Perspective Survey of Trustworthiness of AI in Healthcare
Ojha, J., Presacan, O., Lind, P. G., Monteiro, E., & Yazidi, A. (2025) · ACM Transactions on Computing for Healthcare · doi:10.1145/3716317 · Cited by 41
Counterfactual explanations for misclassified images: How human and machine explanations differ
Delaney, E. D., Pakrashi, A., Greene, D., & Keane, M. T. (2023) · Artificial Intelligence · doi:10.1016/j.artint.2023.103995 · Cited by 43
Deep Learning Uncertainty in Machine Teaching
Sanchez, T., Caramiaux, B., Thiel, P., & Mackay, W. E. (2022) · ACM International Conference on Intelligent User Interfaces (IUI) · doi:10.1145/3490099.3511117 · Cited by 35
“The field of Explainable AI (XAI) explores the role of uncertainty to explainML predictions and shape people’s trust in ML-based decision-making systems [4, 8, 50].”
Bayesian Deep Learning via Subnetwork InferencePreprint
Daxberger, E., Nalisnick, E., Allingham, J. U., Antorán, J., & Hernández-Lobato, J. M. (2020) · UvA-DARE (University of Amsterdam) · doi:10.48550/arxiv.2010.14689 · Cited by 95
“To reliably inform decision making, NNs need to robustly quantify the uncertainty in their predictions (Bhatt et al., 2020).”
BackgroundSelf-citationVeriPlan: Integrating Formal Verification and LLMs into End-User Planning
Lee, C. P., Porfirio, D., Wang, X. J., Zhao, K. C., & Mutlu, B. (2025) · Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3714113 · Cited by 21
“Another usability barrier is users’ difficulty understanding how prompts influence outputs and building accurate mental models of the system’s behavior and the reasoning behind it [7, 101, 107].”
Schrills, T., & Franke, T. (2023) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/3588594 · Cited by 32
“In XAI research explanatory approaches partly refer to the confidence of the model [ 10 , 17 , 86 ] for a certain result or even to meta-information about the model [ 75 ].”
MethodCalibrated explanations: With uncertainty information and counterfactuals
Löfström, H., Löfström, T., Johansson, U., & Sönströd, C. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.123154 · Cited by 44
Transparency improves the accuracy of automation use, but automation confidence information does not
Tatasciore, M., Strickland, L. J. G., & Loft, S. D. (2024) · Cognitive Research Principles and Implications · doi:10.1186/s41235-024-00599-x · Cited by 28
“However, one drawback of using predefined categories is that users may not be aware of or misinterpret the threshold criteria of the categories (Bhatt et al., 2021).”
InfluentialHuman Uncertainty in Concept-Based AI Systems
Collins, K. M., Barker, M., Zarlenga, M. E., Raman, N. J., Bhatt, U., Jamnik, M., Sucholutsky, I., Weller, A., & Dvijotham, K. (2023) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1145/3600211.3604692 · Cited by 58
“Just as machines “knowing when they don’t know” has been emphasized for reliability [2, 21, 35, 43], we emphasize empowering humans to express when they do not know as a way to improve trustworthy deployment and outcomes.”
BackgroundSelf-citationOut-of-Distribution Detection Algorithms for Robust Insect Classification
Saadati, M., Balu, A., Chiranjeevi, S., Jubery, T. Z., Singh, A. K., Sarkar, S., Singh, A., & Ganapathysubramanian, B. (2024) · Plant Phenomics · doi:10.34133/plantphenomics.0170 · Cited by 25
“Increasing the trustworthiness of ML models is an active area of research with several promising approaches [25, 26, 27].”
BackgroundUsing AntiPatterns to avoid MLOps MistakesPreprint
Muralidhar, N., Muthiah, S., Butler, P., Jain, M., Yu, Y., Burne, K., Li, W., Jones, D., Arunachalam, P., McCormick, H. W., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.00079 · Cited by 34
“Recent work [4] shows that in addition to explainability, conveying uncertainty can be a significant contributor to ensuring trust in ML pipelines.”
BackgroundXu, Z., Song, T., & Lee, Y.-C. (2025) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2025.103455 · Cited by 18
Uncertainty Quantification 360: A Hands-on Tutorial
Ghosh, S., Liao, Q. V., Ramamurthy, K. N., Navrátil, J., Sattigeri, P., Varshney, K. R., & Zhang, Y. (2022) · Joint International Conference on Data Science & Management of Data (CODS-COMAD) · doi:10.1145/3493700.3493767 · Cited by 17
“Hence, there is a renewed interest in improving the reliability and transparency of AI models [1].”
BackgroundSelf-citationAI Accountability: Approaches, Affecting Factors, and Challenges
Raja, A. K., & Zhou, J. J. (2023) · Computer · doi:10.1109/mc.2023.3238390 · Cited by 35
Question-Driven Design Process for Explainable AI User Experiences
Liao, Q. V., Pribić, M., Han, J., Miller, S., & Sow, D. (2021) · arXiv · doi:10.48550/arxiv.2104.03483 · Cited by 18
“For model uncertainty, both the properties of the underlying model and user characteristics should guide the selection of both the algorithm to quantify uncertainty and the method to communicate uncertainty [16].”
Self-citationUncertainty-Aware Health Diagnostics via Class-Balanced Evidential Deep Learning
Xia, T., Dang, T., Han, J., Qendro, L., & Mascolo, C. (2024) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2024.3360002 · Cited by 29
“This capabilityisparticularlycrucialinsafety-criticalapplications likehealthdiagnostics[1],[2].”
BackgroundUnpacking Human-AI interactions: From Interaction Primitives to a Design Space
Tsiakas, K., & Murray-Rust, D. (2024) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/3664522 · Cited by 30
On (assessing) the fairness of risk score models
Petersen, E., Ganz, M., Holm, S., & Feragen, A. (2023) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3593013.3594045 · Cited by 37
“Risk scores communicate uncertainty about the potential outcomes to users [7], thus enabling meaningful human oversight and decision-making as required for high-risk AI applications, e.”
BackgroundLi, D., Qiu, T., Chen, S., Li, Q., & Xu, S. (2021) · Annual Computer Security Applications Conference · doi:10.1145/3485832.3485916 · Cited by 30
“Statement There are two kinds of uncertainty associated with machine learning: epistemicvs. aleatoric [39]. The uncertainty tells us about which region of the input space is not perceived by a model [5]. That is, a data sample in the dense region will get a low epistemic uncertainty and get a high epistemic uncertaint…”
BackgroundBusari, I., Sahoo, D., Sudheer, K. P., Harmel, R. D., Privette, C. V., Schlautman, M. A., & Sawyer, C. (2024) · Ecological Informatics · doi:10.1016/j.ecoinf.2024.102735 · Cited by 27
Designing the future smart campus: integrating key elements to enhance user experience
Elbertsen, M., Kok, H., & Salimi, N. (2025) · Journal of Science and Technology Policy Management · doi:10.1108/jstpm-10-2024-0414 · Cited by 26
Training on the Test Set: Mapping the System-Problem Space in AI
Hernández‐Orallo, J., Schellaert, W., & Martínez‐Plumed, F. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i11.21487 · Cited by 32
“Finally, there is much lacking on the explain-ability front as well; we would like to know where a system might fail (city centre in the example) leading to possibly also knowing why (Bhatt et al. 2021).”
BackgroundPapaoikonomou, A., Wingate, J., Verma, V., Durrant, A., Ioannou, G., Papagiannis, T., Yu, M., Alexandridis, G., Dokhane, A., Leontidis, G., et al. (2022) · Annals of Nuclear Energy · doi:10.1016/j.anucene.2022.109373 · Cited by 17
“Lastly, the interpretability of our AI systems via uncertainty measurements (Bhatt et al., 2021), explainable reasoning (Linardatos et al.”
MethodHernández, A. D., Owen, R., Nielsen, D. S., & McConville, R. (2023) · Journal of Responsible Innovation · doi:10.1080/23299460.2023.2222514 · Cited by 17
“ML researchers have tried to understand and measure uncertainty in different ways [2,35].”
BackgroundEstablishing Appropriate Trust in AI through Transparency and Explainability
Kim, S. S. Y. (2024) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3613905.3638184 · Cited by 28
“As a first step, I have been examining the impact of LLMs’ uncertainty expression, a form of transparency [4], on user reliance and trust.”
A Survey on Evidential Deep Learning For Single-Pass Uncertainty Estimation.Preprint
Ulmer, D. (2021) · arXiv (Cornell University) · Cited by 25
Reliable and Trustworthy Machine Learning for Health Using Dataset Shift DetectionPreprint
Park, C., Awadalla, A., Kohno, T., & Patel, S. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.14019 · Cited by 51
“[8] proposed to leverage uncertainty for users making decision and placing trust in machine learning models.”
BackgroundMajlatow, M., Shakil, F. A., Emrich, A., & Mehdiyev, N. (2025) · Applied Sciences · doi:10.3390/app15147925 · Cited by 21
Responsible model deployment via model-agnostic uncertainty learning
Lahoti, P., Gummadi, K. P., & Weikum, G. (2022) · Machine Learning · doi:10.1007/s10994-022-06248-y · Cited by 20
“Trustworthy ML needs models and tools for detecting such failure risks and analyzing the underlying sources of uncertainty (Bhatt et al. 2021).”
BackgroundThe AI-DEC: A Card-based Design Method for User-centered AI Explanations
Lee, C. P., Lee, M. K., & Mutlu, B. (2024) · Designing Interactive Systems Conference · doi:10.1145/3643834.3661576 · Cited by 22
“…theory into AI explanations [134], emphasizing the consideration of situational factors in AI explanations [36], utilizing uncertainty for transparency [16], supporting end-user participation in AI decision-making [75], and providing guidelines for communicating AI trust-worthiness [78].”
BackgroundEllis, C. A., Miller, R. L., & Calhoun, V. D. (2023) · Informatics in Medicine Unlocked · doi:10.1016/j.imu.2023.101176 · Cited by 18
“While there are multiple dimensions to transparency [12], there is little or no available literature on the subject of integrating two of those dimensions - model confidence estimates [13], [14] and model explainability [2][15] – into the same model for neuroimaging classification.”
BackgroundCooperative Learning for Personalized Context-Aware Pain Assessment From Wearable Data
Uddin, M. T., Zamzmi, G., & Canavan, S. J. (2023) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2023.3294903 · Cited by 15
“Furthermore, Q from the proposed COOL framework can be used to communicate the trust behind the inference made by the PAM as it quantifies uncertainty regarding the inference, which is essential for reliability [11], [25], [30], [33], [39].”
MethodCombining uncertainty information with AI recommendations supports calibration with domain knowledge
Subramanian, H. V., Canfield, C., Shank, D. B., & Kinnison, M. (2023) · Journal of Risk Research · doi:10.1080/13669877.2023.2259406 · Cited by 17
“Regardless of the specific metric, uncertainty can be communicated via text, numbers, and visuals as appropriate depending on task and individual characteristics (Bhatt et al. 2021; van der Bles et al. 2019).”
BackgroundA unified and practical user-centric framework for explainable artificial intelligence
Kaplan, S., Uusitalo, H. M. T., & Lensu, L. T. (2023) · Knowledge-Based Systems · doi:10.1016/j.knosys.2023.111107 · Cited by 31
Calibrated explanations for regression
Löfström, T., Löfström, H., Johansson, U., Sönströd, C., & Matela, R. (2025) · Machine Learning · doi:10.1007/s10994-024-06642-8 · Cited by 22
“Therefore, a reliable explanation must have feature weights that correspond accurately to the actual impact on the estimates to correctly reflect the model’s behavior [19].”
Uncertainty-aware visual analytics: scope, opportunities, and challenges
Maack, R. G. C., Scheuermann, G., Hagen, H., Peñaloza, J. T. H., & Gillmann, C. (2022) · The Visual Computer · doi:10.1007/s00371-022-02733-6 · Cited by 25
Delaney, E. D., Greene, D., & Keane, M. T. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.09734 · Cited by 39
“While a rich seam of research exists in uncertainty estimation, much of this work is relatively untapped, albeit extremely promising in the context of generating and evaluating explanations (Bhatt et al., 2020).”
BackgroundMethodInfluentialFair Machine Learning in Healthcare: A Survey
Feng, Q., Du, M., Zou, N., & Hu, X. (2025) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2024.3361836 · Cited by 12
“The measurement and communication of uncertainties are crucial for identifying potential unfairness in model predictions [16].”
Monitoring Model Deterioration with Explainable Uncertainty Estimation via Non-parametric Bootstrap
Mougan, C., & Nielsen, D. S. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i12.26755 · Cited by 25
“Such explainability techniques can provide algorithmic transparency to stakeholders and to the ML engineering team (Mougan, Kanellos, and Gottron 2021; Bhatt et al. 2021; Koh and Liang 2017; Ribeiro, Singh, and Guestrin 2016; Sundararajan, Taly, and Yan 2017).”
BackgroundMehdiyev, N., Majlatow, M., & Fettke, P. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-63800-8_21 · Cited by 16
Mehdiyev, N., Majlatow, M., & Fettke, P. (2024) · Data & Knowledge Engineering · doi:10.1016/j.datak.2024.102402 · Cited by 21
Hopkins, A., & Booth, S. (2021) · DSpace@MIT (Massachusetts Institute of Technology) · Cited by 12
Design, Development, and Deployment of Context-Adaptive AI Systems for Enhanced User Adoption
Lee, C. P. (2024) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3613905.3638195 · Cited by 18
“Efforts are directed towards enhancing transparency in technical models [28, 37], encompassing frameworks for documenting models and datasets [2, 4, 7], methods for elucidating model outputs [17, 31, 34], and strategies for effectively communicating uncertainty [5, 8].”
BackgroundZhu, Z., Tan, Y., Yamashita, N., Lee, Y., & Zhang, R. (2025) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713116 · Cited by 22
“Although research has found that expressing uncertainty can enhance people’s trust toward AI [ 6], it has also been found that AI agents requesting help amplify people’s perceptions of the agents’ incompetence and reduce trust [46].”
Brachman, M., Pan, Q., Do, H. J., Dugan, C., Chaudhary, A., Johnson, J. M., Rai, P., Chakraborti, T., Gschwind, T., Laredo, J. A., et al. (2023) · Proceedings of the 28th International Conference on Intelligent User Interfaces · doi:10.1145/3581641.3584088 · Cited by 11
“Providing the confidence of AI models is also a common and beneficial explanation method for users in generative AI systems [115, 132], as well as decision-based machine learning systems [18].”
MethodKnowles, B., D’Cruz, J., Richards, J. T., & Varshney, K. R. (2023) · Communications of the ACM · doi:10.1145/3587035 · Cited by 17
Using system dynamics to support a functional exercise for pandemic preparedness and response
Green, C., Beishuizen, B. H. H., Stein, M. L., Bleeker‐Rovers, C. P., Tostmann, A., de Jong, D. L. K., Houareau, C., Perseke, K., Spieker, C., Grote, U., et al. (2024) · System Dynamics Review · doi:10.1002/sdr.1786 · Cited by 12
How Displaying AI Confidence Affects Reliance and Hybrid Human-AI Performance
Tejeda, H., Kumar, A., & Mark Steyvers, B. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230087 · Cited by 14
Ulmer, D., Frellsen, J., & Hardmeier, C. (2022) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2022.findings-emnlp.198 · Cited by 46
Metadata
- Authors
- Bhatt, U., Antorán, J., Zhang, Y., Liao, Q. V., Sattigeri, P., Fogliato, R., Melançon, G., Krishnan, R., Stanley, J., Tickoo, O., Nachman, L., Chunara, R., Srikumar, M., Weller, A., & Xiang, A.
- Affiliations
- University of Cambridge (GB)
- IBM (United States) (US)
- Carnegie Mellon University (US)
- Concordia University (CA)
- Intel (United States) (US)
- New York University (US)
- Partnership on AI, San Francisco, CA, USA
- Sony AI, Seattle, WA, USA
- ORCID
- Venue
- Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society
- Publisher
- ACM
- Published
- 2021-07-21
- Pages
- 401-413
- ISSN
- 3065-8365 (ISSN-L)
- OpenAlex
- W3183398589
- Semantic Scholar
- 3973e0fab69a00f5ed6a81ca408f60c420fa6e61
- Cited by
- 300
- References
- 116
- Citations by year
- 2026: 59 · 2025: 62 · 2024: 65 · 2023: 46 · 2022: 48 · 2021: 18
- FWCI
- 16.64
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Proceedings of the AAAI/ACM Conference on AI Ethics and Society (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- Cambridge University Engineering Department Publications Database (submittedVersion)
- arXiv (Cornell University)
- Funders
- Microsoft Research
- Northwestern University
- Alan Turing Institute (EP/N510129/1., TU/B/000074, N510129)
- DeepMind
- Leverhulme Trust
- Engineering and Physical Sciences Research Council (N510129, EP/N510129/, EP/N510129/1, TU/B/000074, 2275741, EPSRC grant EP/N510129/1, EP/V025279/1)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
- Keywords
- algorithmic transparency, uncertainty estimation, uncertainty communication, explainability, model fairness, trustworthy systems, decision-making, uncertainty visualization, human-computer interaction, machine learning pipelines, stakeholder awareness
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 2026-10-24, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-24 (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.