mlsanityData to 7 Oct 2026

Reducing conversational agents' overconfidence through linguistic calibration

arXiv 2012.14983 ↗First posted 30 Dec 2020cs.CL cs.AI cs.LGcited by 209

Not on the map. Why?

Abstract (arXiv)

While improving neural dialogue agents' factual accuracy is the object of much research, another important aspect of communication, less studied in the setting of neural dialogue, is transparency about ignorance. In this work, we analyze to what extent state-of-the-art chit-chat models are linguistically calibrated in the sense that their verbalized expression of doubt (or confidence) matches the likelihood that the model's responses are factually incorrect (or correct). We find that these models are poorly calibrated, yet we show that likelihood of correctness can accurately be predicted. By incorporating such metacognitive features into the training of a controllable generation model, we obtain a dialogue agent with greatly improved linguistic calibration. While improving neural dialogue agents' factual accuracy is the object of much research, another important aspect of communication, less studied in the setting of neural dialogue, is transparency about ignorance. In this work, we analyze to what extent state-of-the-art chit-chat models are linguistically calibrated in the sense that their verbalized expression of doubt (or confidence) matches the likelihood that the model's responses are factually incorrect (or correct). We find that these models are poorly calibrated, yet we show that likelihood of correctness can accurately be predicted. By incorporating such metacognitive features into the training of a controllable generation model, we obtain a dialogue agent with greatly improved linguistic calibration.

Key facts. Cited by 209 papers in the map (data through 7 Oct 2026). First posted on arXiv 30 Dec 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 6; 2023: 33; 2024: 70; 2025: 61; 2026: 39.

209 papers.

Sort:

By area:

  1. 2023Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMscited by 946
  2. 2023Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedbackcited by 788
  3. 2022Teaching Models to Express Their Uncertainty in Wordscited by 731
  4. 2023Chain-of-Verification Reduces Hallucination in Large Language Modelscited by 317
  5. 2023Robots That Ask For Help: Uncertainty Alignment for Large Language Model Plannerscited by 298
  6. 2023ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMscited by 297
  7. 2023Generating with Confidence: Uncertainty Quantification for Black-box Large Language Modelscited by 286
  8. 2022Prompting GPT-3 To Be Reliablecited by 258
  9. 2023A Survey of Confidence Estimation and Calibration in Large Language Modelscited by 253
  10. 2024Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMscited by 246
  11. 2025Why Language Models Hallucinatecited by 205
  12. 2022The Unreliability of Explanations in Few-shot Prompting for Textual Reasoningcited by 169
38 more
  1. 2023Head-to-Tail: How Knowledgeable are Large Language Models (LLMs)? A.K.A. Will LLMs Replace Knowledge Graphs?cited by 167
  2. 2023Conformal Language Modelingcited by 163
  3. 2024Kernel Language Entropy: Fine-grained Uncertainty Quantification for LLMs from Semantic Similaritiescited by 156
  4. 2025Reasoning Models Know When They're Right: Probing Hidden States for Self-Verificationcited by 154
  5. 2023Evaluating Correctness and Faithfulness of Instruction-Following Models for Question Answeringcited by 150
  6. 2023Inspecting and Editing Knowledge Representations in Language Modelscited by 140
  7. 2024To Believe or Not to Believe Your LLMcited by 120
  8. 2024Know Your Limits: A Survey of Abstention in Large Language Modelscited by 116
  9. 2024Language Model Cascades: Token-level uncertainty and beyondcited by 115
  10. 2024A Survey on Uncertainty Quantification of Large Language Models: Taxonomy, Open Research Challenges, and Future Directionscited by 115
  11. 2023Decomposing Uncertainty for Large Language Models through Input Clarification Ensemblingcited by 113
  12. 2022FaithDial: A Faithful Benchmark for Information-Seeking Dialoguecited by 108
  13. 2025Beyond Binary Rewards: Training LMs to Reason About Their Uncertaintycited by 107
  14. 2024Relying on the Unreliable: The Impact of Language Models' Reluctance to Express Uncertaintycited by 106
  15. 2024The Art of Saying No: Contextual Noncompliance in Language Modelscited by 100
  16. 2025Uncertainty Quantification and Confidence Calibration in Large Language Models: A Surveycited by 93
  17. 2024On Verbalized Confidence Scores for LLMscited by 83
  18. 2024LUQ: Long-text Uncertainty Quantification for LLMscited by 79
  19. 2023A Survey on Uncertainty Quantification Methods for Deep Learningcited by 78
  20. 2024Explore until Confident: Efficient Exploration for Embodied Question Answeringcited by 74
  21. 2024Uncertainty Estimation and Quantification for LLMs: A Simple Supervised Approachcited by 73
  22. 2024Calibrating Large Language Models with Sample Consistencycited by 72
  23. 2023Do Androids Know They're Only Dreaming of Electric Sheep?cited by 67
  24. 2024"I'm Not Sure, But...": Examining the Impact of Large Language Models' Uncertainty Expression on User Reliance and Trustcited by 67
  25. 2023Uncertainty in Natural Language Generation: From Theory to Applicationscited by 61
  26. 2024Calibrating Language Models with Adaptive Temperature Scalingcited by 57
  27. 2024Mitigating LLM Hallucinations via Conformal Abstentioncited by 56
  28. 2024Latent Space Chain-of-Embedding Enables Output-free LLM Self-Evaluationcited by 56
  29. 2023Knowing What LLMs DO NOT Know: A Simple Yet Effective Self-Detection Methodcited by 51
  30. 2024Towards Human-centered Proactive Conversational Agentscited by 48
  31. 2025Rewarding Doubt: A Reinforcement Learning Approach to Calibrated Confidence Expression of Large Language Modelscited by 48
  32. 2025Calibrating Verbal Uncertainty as a Linear Feature to Reduce Hallucinationscited by 48
  33. 2022RHO (ρ): Reducing Hallucination in Open-domain Dialogues with Knowledge Groundingcited by 47
  34. 2025Creating General User Models from Computer Usecited by 44
  35. 2023Prompting is not a substitute for probability measurements in large language modelscited by 43
  36. 2025Mind the Confidence Gap: Overconfidence, Calibration, and Distractor Effects in Large Language Modelscited by 39
  37. 2024Thermometer: Towards Universal Calibration for Large Language Modelscited by 38
  38. 2024Uncertainty quantification in fine-tuned LLMs using LoRA ensemblescited by 35

50 most cited of 209 All 209 (JSON)