mlsanityData to 7 Oct 2026

Stop Measuring Calibration When Humans Disagree

arXiv 2210.16133 ↗First posted 28 Oct 2022cs.CL cs.AI cs.LGcited by 70

Not on the map. Why?

Abstract (arXiv)

Calibration is a popular framework to evaluate whether a classifier knows when it does not know - i.e., its predictive probabilities are a good indication of how likely a prediction is to be correct. Correctness is commonly estimated against the human majority class. Recently, calibration to human majority has been measured on tasks where humans inherently disagree about which class applies. We show that measuring calibration to human majority given inherent disagreements is theoretically problematic, demonstrate this empirically on the ChaosNLI dataset, and derive several instance-level measures of calibration that capture key statistical properties of human judgements - class frequency, ranking and entropy.

Key facts. Cited by 70 papers in the map (data through 7 Oct 2026). Builds on 21 of its 47 references that are in the map. First posted on arXiv 28 Oct 2022.

Builds on

21 of 47 references.

References · 21

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
  3. 2017On Calibration of Modern Neural Networkscited by 4,860
  4. 2017A Broad-Coverage Challenge Corpus for Sentence Understanding through Inferencecited by 2,665
  5. 2015A large annotated corpus for learning natural language inferencecited by 1,837
  6. 2020A Primer in BERTology: What we know about how BERT workscited by 807
  7. 2019Measuring Calibration in Deep Learningcited by 443
  8. 2020How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answeringcited by 439
  9. 2019Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibrationcited by 359
  10. 2020Calibration of Pre-trained Transformerscited by 307
  11. 2019Abductive Commonsense Reasoningcited by 300
  12. 2019Evaluating model calibration in classificationcited by 165
9 more
  1. 2020What Can We Learn from Collective Human Opinions on Natural Language Inference Data?cited by 148
  2. 2021On Releasing Annotator-Level Labels and Information in Datasetscited by 123
  3. 2020Calibration of Neural Networks using Splinescited by 97
  4. 2019Calibration tests in multi-class classification: A unifying frameworkcited by 79
  5. 2022Annotation Error Detection: Analyzing the Past and Present for a More Coherent Futurecited by 45
  6. 2019Uncertain Natural Language Inferencecited by 35
  7. 2021Distributed NLI: Learning to Predict Human Opinion Distributions for Language Reasoningcited by 31
  8. 2021Learning with Different Amounts of Annotation: From Zero to Many Labelscited by 28
  9. 2021Embracing Ambiguity: Shifting the Training Target of NLI Modelscited by 22

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 2; 2023: 11; 2024: 22; 2025: 17; 2026: 18.

70 papers.

Sort:

By area:

  1. 2023A Survey of Confidence Estimation and Calibration in Large Language Modelscited by 253
  2. 2022The 'Problem' of Human Label Variation: On Ground Truth in Data, Modeling and Evaluationcited by 133
  3. 2023Value Kaleidoscope: Engaging AI with Pluralistic Human Values, Rights, and Dutiescited by 97
  4. 2025Uncertainty Quantification and Confidence Calibration in Large Language Models: A Surveycited by 93
  5. 2024LLM-Rubric: A Multidimensional, Calibrated Approach to Automated Evaluation of Natural Language Textscited by 81
  6. 2023Uncertainty in Natural Language Generation: From Theory to Applicationscited by 61
  7. 2024Conformal Prediction for Natural Language Processing: A Surveycited by 52
  8. 2023What Comes Next? Evaluating Uncertainty in Neural Text Generators Against Human Production Variabilitycited by 51
  9. 2025Specializing Large Language Models to Simulate Survey Response Distributions for Global Populationscited by 50
  10. 2024The Perspectivist Paradigm Shift: Assumptions and Challenges of Capturing Human Labelscited by 37
  11. 2023Can Large Language Models Capture Dissenting Human Voices?cited by 33
  12. 2024The Effectiveness of LLMs as Annotators: A Comparative Overview and Empirical Analysis of Direct Representationcited by 29
38 more
  1. 2025SemEval-2025 Task 3: Mu-SHROOM, the Multilingual Shared Task on Hallucinations and Related Observable Overgeneration Mistakescited by 27
  2. 2023It Takes Two to Tango: Navigating Conceptualizations of NLP Tasks and Measurements of Performancecited by 25
  3. 2023A Study on the Calibration of In-context Learningcited by 21
  4. 2025Understanding Model Calibration -- A gentle introduction and visual exploration of calibration and the expected calibration error (ECE)cited by 20
  5. 2024Conformalized Credal Regions for Classification with Ambiguous Ground Truthcited by 19
  6. 2024Annotator-Centric Active Learning for Subjective NLP Taskscited by 18
  7. 2023Don't Waste a Single Annotation: Improving Single-Label Classifiers Through Soft Labelscited by 17
  8. 2024"Seeing the Big through the Small": Can LLMs Approximate Human Judgment Distributions on NLI from a Few Explanations?cited by 16
  9. 2024Interpreting Predictive Probabilities: Model Confidence or Human Label Variation?cited by 15
  10. 2025Do LLMs Give Psychometrically Plausible Responses in Educational Assessments?cited by 14
  11. 2024Crowd-Calibrator: Can Annotator Disagreement Inform Calibration in Subjective Tasks?cited by 13
  12. 2023Expectations over Unspoken Alternatives Predict Pragmatic Inferencescited by 9
  13. 2024Through the Lens of Split Vote: Exploring Disagreement, Difficulty and Calibration in Legal Case Outcome Classificationcited by 9
  14. 2024"I Never Said That": A dataset, taxonomy and baselines on response clarity classificationcited by 9
  15. 2024Understanding the Impact of Confidence in Retrieval Augmented Generation: A Case Study in the Medical Domaincited by 9
  16. 2025Always Tell Me The Odds: Fine-grained Conditional Probability Estimationcited by 9
  17. 2024A Rose by Any Other Name: LLM-Generated Explanations Are Good Proxies for Human Explanations to Collect Label Distributions on NLIcited by 8
  18. 2023Evaluating AI systems under uncertain ground truth: a case study in dermatologycited by 7
  19. 2024Mind the Uncertainty in Human Disagreement: Evaluating Discrepancies between Model Predictions and Human Responses in VQAcited by 6
  20. 2023Understanding and Predicting Human Label Variation in Natural Language Inference through Explanationcited by 5
  21. 2024Variational Low-Rank Adaptation Using IVONcited by 5
  22. 2025Improving LoRA with Variational Learningcited by 5
  23. 2026SemEval-2026 Task 6: CLARITY -- Unmasking Political Question Evasionscited by 5
  24. 2024Predict the Next Word: Humans exhibit uncertainty in this task and language models _____cited by 4
  25. 2024Confidence Estimation for LLM-Based Dialogue State Trackingcited by 4
  26. 2024Understanding The Effect Of Temperature On Alignment With Human Opinionscited by 4
  27. 2025Unveiling the "Fairness Seesaw": Discovering and Mitigating Gender and Race Bias in Vision-Language Modelscited by 4
  28. 2026LLMs Capture Emotion Labels, Not Emotion Uncertainty: Distributional Analysis and Calibration of Human-LLM Judgment Gapscited by 4
  29. 2022Aggregating Soft Labels from Crowd Annotations Improves Uncertainty Estimation Under Distribution Shiftcited by 3
  30. 2024A Dataset for Physical and Abstract Plausibility and Sources of Human Disagreementcited by 3
  31. 2024A Framework for Evaluating LLMs Under Task Indeterminacycited by 3
  32. 2025Comparing Uncertainty Measurement and Mitigation Methods for Large Language Models: A Systematic Reviewcited by 3
  33. 2025Judging with Confidence: Calibrating Autoraters to Preference Distributionscited by 3
  34. 2025SIMBA UQ: Similarity-Based Aggregation for Uncertainty Quantification in Large Language Modelscited by 2
  35. 2026Uncertainty Decomposition via Cyclical SG-MCMC and Soft-label Learning for Subjective NLPcited by 2
  36. 2024Efficient Nearest Neighbor based Uncertainty Estimation for Natural Language Processing Taskscited by 1
  37. 2024Exploring the Influence of Label Aggregation on Minority Voices: Implications for Dataset Bias and Model Trainingcited by 1
  38. 2025Subjective Logic Encodingscited by 1

50 most cited of 70 All 70 (JSON)