5 papers
Sharp feature-learning transitions and Bayes-optimal neural scaling laws in extensive-width networks
Minh-Toan Nguyen, Jean Barbier
We study the information-theoretic limits of learning a one-hidden-layer teacher network with hierarchical features from noisy queries, in the context of knowledge transfer to a sm…
Med-StepBench: A Hierarchical Reasoning Framework for Evaluating Hallucinations in Medical Vision-Language Models
Minh Khoi Nguyen, Dai Lam Le, Amir Reza Jafari +8
Large vision-language models (VLMs) demonstrate strong performance in medical image understanding, but frequently generate clinically plausible yet incorrect statements, raising si…
Statistical physics of deep learning: Optimal learning of a multi-layer perceptron near interpolation
Jean Barbier, Francesco Camilli, Minh-Toan Nguyen +2
For four decades statistical physics has been providing a framework to analyse neural networks. A long-standing question remained on its capacity to tackle deep learning models cap…
Statistical mechanics of extensive-width Bayesian neural networks near interpolation
Jean Barbier, Francesco Camilli, Minh-Toan Nguyen +2
For three decades statistical mechanics has been providing a framework to analyse neural networks. However, the theoretically tractable models, e.g., perceptrons, random features m…
Optimal generalisation and learning transition in extensive-width shallow neural networks near interpolation
Jean Barbier, Francesco Camilli, Minh-Toan Nguyen +2
We consider a teacher-student model of supervised learning with a fully-trained two-layer neural network whose width and input dimension are large and proportional. We prov…