One Model, Two Precisions: Does Disagreement Between Quantized Copies Tell You When a Small Model Is Wrong?
cs.AI
Submitted: 2026-11-12
Updated: 2026-11-12
Terminology
Sources
- When Quantization Affects Confidence of Large Language Models?
- The Illusion of Equivalency: Statistical Characterization of Quantization Effects in LLMs
- Accuracy is Not All You Need
- How Divergence Becomes Decision Flips in Compressed Language Models
- The Illusion of Certainty: Uncertainty Quantification for LLMs Fails under Ambiguity
- The Role of Ambiguity in Error Prediction via Uncertainty Quantification
- Revisiting Uncertainty Estimation and Calibration of Large Language Models
- Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
- Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
- A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
- On Calibration of Modern Neural Networks
- Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift
- Deep Ensembles Work, But Are They Necessary?
- BatchEnsemble: An Alternative Approach to Efficient Ensemble and Lifelong Learning
- Packed-Ensembles for Efficient Uncertainty Estimation
- Quantifying the Uncertainty of Foundation Models with Singular Value Ensembles
- Training-Free Bayesianization for Low-Rank Adapters of Large Language Models
- Selective Classification for Deep Neural Networks
- SelectiveNet: A Deep Neural Network with an Integrated Reject Option
- Overcoming Common Flaws in the Evaluation of Selective Classification Systems
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- 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
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents