When Do Supervised UQ Ensembles Improve LLM Hallucination Detection? A Robustness Study
cs.LG, cs.AI, cs.CL
Submitted: 2026-08-25
Updated: 2026-09-08
Comments: Accepted to AACL 2026 (Findings)
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Big-Math: A Large-Scale, High-Quality Math Dataset for Reinforcement Learning in Language Models
- Reconsidering LLM Uncertainty Estimation Methods in the Wild
- Functional Entropy: Predicting Functional Correctness in LLM-Generated Code with Uncertainty Quantification
- Language Models (Mostly) Know What They Know
- CodeBLEU: a Method for Automatic Evaluation of Code Synthesis
- EAGer: Entropy-Aware GEneRation for Adaptive Inference-Time Scaling
- Assessing Correctness in LLM-Based Code Generation via Uncertainty Estimation
- Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback
- Replacing Judges with Juries: Evaluating LLM Generations with a Panel of Diverse Models
- Measuring short-form factuality in large language models
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks