mlsanityData to 14 Oct 2026

Large language model validity via enhanced conformal prediction methods

arXiv 2406.09714 ↗First posted 14 Jun 2024stat.ML cs.LG stat.MEcited by 92

Not on the map. Why?

Abstract (arXiv)

We develop new conformal inference methods for obtaining validity guarantees on the output of large language models (LLMs). Prior work in conformal language modeling identifies a subset of the text that satisfies a high-probability guarantee of correctness. These methods work by filtering claims from the LLM's original response if a scoring function evaluated on the claim fails to exceed a threshold calibrated via split conformal prediction. Existing methods in this area suffer from two deficiencies. First, the guarantee stated is not conditionally valid. The trustworthiness of the filtering step may vary based on the topic of the response. Second, because the scoring function is imperfect, the filtering step can remove many valuable and accurate claims. We address both of these challenges via two new conformal methods. First, we generalize the conditional conformal procedure of Gibbs et al. (2023) in order to adaptively issue weaker guarantees when they are required to preserve the utility of the output. Second, we show how to systematically improve the quality of the scoring function via a novel algorithm for differentiating through the conditional conformal procedure. We demonstrate the efficacy of our approach on biography and medical question-answering datasets.

Key facts. Cited by 92 papers in the map (data through 14 Oct 2026). Builds on 20 of its 36 references that are in the map. First posted on arXiv 14 Jun 2024.

Builds on

20 of 36 references.

References · 20

  1. 2022Language Models (Mostly) Know What They Knowcited by 1,856
  2. 2022Large Language Models Encode Clinical Knowledgecited by 1,830
  3. 2023FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generationcited by 1,282
  4. 2023SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Modelscited by 810
  5. 2023The Internal State of an LLM Knows When It's Lyingcited by 659
  6. 2022Conformal Risk Controlcited by 299
  7. 2021Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Controlcited by 271
  8. 2023A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of LLMs by Validating Low-Confidence Generationcited by 206
  9. 2023Conformal Language Modelingcited by 170
  10. 2023Conformal Prediction With Conditional Guaranteescited by 152
  11. 2023Conformal Prediction with Large Language Models for Multi-Choice Question Answeringcited by 139
  12. 2024Language Models with Conformal Factuality Guaranteescited by 131
8 more
  1. 2024Benchmarking LLMs via Uncertainty Quantificationcited by 126
  2. 2021Learning Optimal Conformal Classifierscited by 112
  3. 2024Multicalibration for Confidence Scoring in LLMscited by 46
  4. 2022Low-Degree Multicalibrationcited by 45
  5. 2024K-QA: A Real-World Medical Q&A Benchmarkcited by 31
  6. 2024C-RAG: Certified Generation Risks for Retrieval-Augmented Language Modelscited by 27
  7. 2024OLAPH: Improving Factuality in Biomedical Long-form Question Answeringcited by 20
  8. 2024Benchmarking Llama2, Mistral, Gemma and GPT for Factuality, Toxicity, Bias and Propensity for Hallucinationscited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 11; 2025: 40; 2026: 41.

92 papers.

Sort:

By area:

  1. 2025Uncertainty Quantification and Confidence Calibration in Large Language Models: A Surveycited by 95
  2. 2024Knowledge Boundary of Large Language Models: A Surveycited by 40
  3. 2024Length Optimization in Conformal Predictioncited by 39
  4. 2025From Calibration to Collaboration: LLM Uncertainty Quantification Should Be More Human-Centeredcited by 21
  5. 2025Conformal Linguistic Calibration: Trading-off between Factuality and Specificitycited by 20
  6. 2025Steering Dialogue Dynamics for Robustness against Multi-turn Jailbreaking Attackscited by 20
  7. 2025CP-Router: An Uncertainty-Aware Router Between LLM and LRMcited by 20
  8. 2026Uncertainty Quantification in LLM Agents: Foundations, Emerging Challenges, and Opportunitiescited by 20
  9. 2024Sample then Identify: A General Framework for Risk Control and Assessment in Multimodal Large Language Modelscited by 19
  10. 2024Prune 'n Predict: Optimizing LLM Decision-making with Conformal Predictioncited by 18
  11. 2025E-valuator: Reliable Agent Verifiers with Sequential Hypothesis Testingcited by 17
  12. 2025Backward Conformal Predictioncited by 14
38 more
  1. 2025Conformal Language Model Reasoning with Coherent Factualitycited by 14
  2. 2024Multi-group Uncertainty Quantification for Long-form Text Generationcited by 13
  3. 2025Conformal Arbitrage: Risk-Controlled Balancing of Competing Objectives in Language Modelscited by 12
  4. 2024Conformal Risk Minimization with Variance Reductioncited by 11
  5. 2024Conformal Generative Modeling with Improved Sample Efficiency through Sequential Greedy Filteringcited by 10
  6. 2025Conformal Prediction Sets for Deep Generative Models via Reduction to Conformal Regressioncited by 10
  7. 2025PCS-UQ: Uncertainty Quantification via the Predictability-Computability-Stability Frameworkcited by 9
  8. 2025Lexical Hints of Accuracy in LLM Reasoning Chainscited by 9
  9. 2025Conformal Tail Risk Control for Large Language Model Alignmentcited by 8
  10. 2025Response Quality Assessment for Retrieval-Augmented Generation via Conditional Conformal Factualitycited by 8
  11. 2024C-Adapter: Adapting Deep Classifiers for Efficient Conformal Prediction Setscited by 7
  12. 2025Mirror Online Conformal Prediction with Intermittent Feedbackcited by 7
  13. 2025Conformal Information Pursuit for Interactively Guiding Large Language Modelscited by 7
  14. 2025Human-AI Collaborative Uncertainty Quantificationcited by 7
  15. 2026CORA: Conformal Risk-Controlled Agents for Safeguarded Mobile GUI Automationcited by 7
  16. 2025Conformal Prediction Beyond the Seen: A Missing Mass Perspective for Uncertainty Quantification in Generative Modelscited by 6
  17. 2025Document Summarization with Conformal Importance Guaranteescited by 6
  18. 2025Domain-Shift-Aware Conformal Prediction for Large Language Modelscited by 6
  19. 2026Conformal Policy Controlcited by 6
  20. 2024Spatial-Aware Conformal Prediction for Trustworthy Hyperspectral Image Classificationcited by 5
  21. 2025Exploring the Noise Robustness of Online Conformal Predictioncited by 5
  22. 2025Do Large Language Models (Really) Need Statistical Foundations?cited by 5
  23. 2025Conformal Prediction for Time-series Forecasting with Change Pointscited by 5
  24. 2025Adaptive Coverage Policies in Conformal Predictioncited by 5
  25. 2024Conditional Testing based on Localized Conformal p-valuescited by 4
  26. 2025MCQA-Eval: Efficient Confidence Evaluation in NLG with Gold-Standard Correctness Labelscited by 4
  27. 2025Multi-Condition Conformal Selectioncited by 4
  28. 2025Knowing What You Know Is Not Enough: Large Language Model Confidences Don't Align With Their Actionscited by 4
  29. 2026Online Reasoning Calibration: Test-Time Training Enables Generalizable Conformal LLM Reasoningcited by 4
  30. 2026Online Conformal Prediction with Adversarial Semi-bandit Feedback via Regret Minimizationcited by 4
  31. 2026When Can Conformal Risk Control Certify LLM Outputs? Bounds, Impossibility, and Adaptation for Structured Generationcited by 4
  32. 2026Doomed from the Start: Early Abort of LLM Agent Episodes via a Recall-Controlled Probe Cascadecited by 4
  33. 2025Calibrated Predictive Lower Bounds on Time-to-Unsafe-Sampling in LLMscited by 3
  34. 2025Filtering with Confidence: When Data Augmentation Meets Conformal Predictioncited by 3
  35. 2026Entropy-Tree: Tree-Based Decoding with Entropy-Guided Explorationcited by 3
  36. 2026Unsupervised Confidence Calibration for Reasoning LLMs from a Single Generationcited by 3
  37. 2025Bayesian Modeling of Zero-Shot Classifications for Urban Flood Detectioncited by 2
  38. 2025ConfAgents: A Conformal-Guided Multi-Agent Framework for Cost-Efficient Medical Diagnosiscited by 2

50 most cited of 92 All 92 (JSON)