5 papers
A Function-Space Dichotomy for Compositional Learning: Exponential Sub-Optimality of the Neural Tangent Kernel
Arkaprabha Ganguli, Emil Constantinescu
A persistent empirical observation is that trained neural networks outperform their neural tangent kernel (NTK) limit on tasks with compositional structure, yet a quantitative acco…
Event-Level QCD Inference Framework for Quark-Gluon Imaging
Patrick Barry, Pi-Yueh Chuang, Ian Cloët +3
We introduce and demonstrate an event-level analysis framework for quark-gluon imaging. For a first application we use it for the inference of parton distribution functions from sy…
Disentangled Deep Priors for Bayesian Inverse Problems
Arkaprabha Ganguli, Emil Constantinescu
We propose a structured prior for high-dimensional Bayesian inverse problems based on a disentangled deep generative model whose latent space is partitioned into auxiliary variable…
Uncovering Physical Drivers of Dark Matter Halo Structures with Auxiliary-Variable-Guided Generative Models
Arkaprabha Ganguli, Anirban Samaddar, Florian Kéruzoré +4
Deep generative models (DGMs) compress high-dimensional data but often entangle distinct physical factors in their latent spaces. We present an auxiliary-variable-guided framework…
Enhancing Interpretability in Generative Modeling: Statistically Disentangled Latent Spaces Guided by Generative Factors in Scientific Datasets
Arkaprabha Ganguli, Nesar Ramachandra, Julie Bessac +1
This study addresses the challenge of statistically extracting generative factors from complex, high-dimensional datasets in unsupervised or semi-supervised settings. We investigat…