mlsanityData to 7 Oct 2026

Towards Reliable Truth-Aligned Uncertainty Estimation in Large Language Models

arXiv 2604.00445 ↗First posted 1 Apr 2026cs.AI cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Uncertainty estimation (UE) aims to detect hallucinated outputs of large language models (LLMs) to improve their reliability. However, UE metrics often exhibit unstable performance across configurations, which significantly limits their applicability. In this work, we formalise this phenomenon as proxy failure, since most UE metrics originate from model behaviour, rather than being explicitly grounded in the factual correctness of LLM outputs. With this, we show that UE metrics become non-discriminative precisely in low-information regimes. To alleviate this, we propose Truth AnChoring (TAC), a post-hoc calibration method to remedy UE metrics, by mapping the raw scores to truth-aligned scores. Even with noisy and few-shot supervision, our TAC can support the learning of well-calibrated uncertainty estimates, and presents a practical calibration protocol. Our findings highlight the limitations of treating heuristic UE metrics as direct indicators of truth uncertainty, and position our TAC as a necessary step toward more reliable uncertainty estimation for LLMs. The code repository is available at https://github.com/ponhvoan/TruthAnchor/.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 24 of its 43 references that are in the map. First posted on arXiv 1 Apr 2026.

Builds on

24 of 43 references.

References · 24

  1. 2017On Calibration of Modern Neural Networkscited by 4,860
  2. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,895
  3. 2020DeBERTa: Decoding-enhanced BERT with Disentangled Attentioncited by 2,157
  4. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  5. 2020Energy-based Out-of-distribution Detectioncited by 1,296
  6. 2021DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharingcited by 1,219
  7. 2022When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memoriescited by 1,103
  8. 2023BloombergGPT: A Large Language Model for Financecited by 923
  9. 2023Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedbackcited by 788
  10. 2017Crowdsourcing Multiple Choice Science Questionscited by 742
  11. 2022Discovering Latent Knowledge in Language Models Without Supervisioncited by 733
  12. 2023The Internal State of an LLM Knows When It's Lyingcited by 644
12 more
  1. 2023Capabilities of GPT-4 on Medical Challenge Problemscited by 630
  2. 2022Prompting GPT-3 To Be Reliablecited by 258
  3. 2024Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Mapscited by 140
  4. 2024LLM Internal States Reveal Hallucination Risk Faced With a Querycited by 80
  5. 2024Latent Space Chain-of-Embedding Enables Output-free LLM Self-Evaluationcited by 56
  6. 2024Characterizing Truthfulness in Large Language Model Generations with Local Intrinsic Dimensioncited by 45
  7. 2025Semantic Volume: Quantifying and Detecting both External and Internal Uncertainty in LLMscited by 19
  8. 2025Token-Level Density-Based Uncertainty Quantification Methods for Eliciting Truthfulness of Large Language Modelscited by 18
  9. 2025Too Consistent to Detect: A Study of Self-Consistent Errors in LLMscited by 18
  10. 2025The Illusion of Certainty: Uncertainty Quantification for LLMs Fails under Ambiguitycited by 15
  11. 2025Unsupervised Hallucination Detection by Inspecting Reasoning Processescited by 5
  12. 2025Measuring Aleatoric and Epistemic Uncertainty in LLMs: Empirical Evaluation on ID and OOD QA Taskscited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Revisiting Uncertainty Estimation and Calibration of Large Language Models
  2. 2026Rethinking Uncertainty Evaluation in Large Language Models
  3. 2025Trusted Uncertainty in Large Language Models: A Unified Framework for Confidence Calibration and Risk-Controlled Refusal
  4. 2024Calibrating the Confidence of Large Language Models by Eliciting Fidelity
  5. 2026Complementing Self-Consistency with Cross-Model Disagreement for Uncertainty Quantification
  6. 2025Simple Yet Effective: An Information-Theoretic Approach to Multi-LLM Uncertainty Quantification
  7. 2026ChainUQ: Reasoning Consistency-Aware Uncertainty Quantification for Large Language Models
  8. 2026Sensitivity Uncertainty Alignment in Large Language Models ↗