Towards Reliable Truth-Aligned Uncertainty Estimation in Large Language Models

arXiv:2604.00445 · cs.AI, cs.CL · Submitted 2026-04-01 · Read on arXiv

cs.AI, cs.CL

Submitted: 2026-04-01

Updated: 2026-10-16

Comments: Accepted to EMNLP Findings 2026

Code: https://github.com/ponhvoan/TruthAnchor

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Calibration — 2× in this paper · explained in 9 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 8 episodes

Sources

• GPT-4 Technical Report
• Discovering Latent Knowledge in Language Models Without Supervision
• INSIDE: LLMs' Internal States Retain the Power of Hallucination Detection
• Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Maps
• Training Verifiers to Solve Math Word Problems
• The Llama 3 Herd of Models
• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
• DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing
• Look Before You Leap: An Exploratory Study of Uncertainty Measurement for Large Language Models
• Language Models (Mostly) Know What They Know
• Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generation
• Semantic Volume: Quantifying and Detecting both External and Internal Uncertainty in LLMs
• Generating with Confidence: Uncertainty Quantification for Black-box Large Language Models
• Ministral 3
• Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models
• Uncertainty Estimation in Autoregressive Structured Prediction
• SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models
• Capabilities of GPT-4 on Medical Challenge Problems
• Prompting GPT-3 To Be Reliable
• Gemma 2: Improving Open Language Models at a Practical Size

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems