Image de couverture de livre

A provable control of sensitivity of neural networks through a direct parameterization of the overall bi-Lipschitzness

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Kinoshita, Yuri, Toyoizumi, Taro
Format:	Preprint
Publié:	2024
Sujets:	Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Dataset Distillation Efficiently Encodes Low-Dimensional Representations from Gradient-Based Learning of Non-Linear Tasks
par: Kinoshita, Yuri, et autres
Publié: (2026)

Cortex and subcortex play distinct roles over learning when cortical memory is limited
par: Farrell, Matthew, et autres
Publié: (2026)

Residual connections provably mitigate oversmoothing in graph neural networks
par: Chen, Ziang, et autres
Publié: (2025)

On the expressivity of bi-Lipschitz normalizing flows
par: Verine, Alexandre, et autres
Publié: (2021)

Towards graph neural networks for provably solving convex optimization problems
par: Qian, Chendi, et autres
Publié: (2025)

Principles of Lipschitz continuity in neural networks
par: Luo, Róisín
Publié: (2026)

Controlling Posterior Collapse by an Inverse Lipschitz Constraint on the Decoder Network
par: Kinoshita, Yuri, et autres
Publié: (2023)

Efficient and provably convergent end-to-end training of deep neural networks with linear constraints
par: Yang, Zonglin, et autres
Publié: (2026)

Lipschitz-bounded 1D convolutional neural networks using the Cayley transform and the controllability Gramian
par: Pauli, Patricia, et autres
Publié: (2023)

Lipschitz constant estimation for general neural network architectures using control tools
par: Pauli, Patricia, et autres
Publié: (2024)

Nonparametric regression using over-parameterized shallow ReLU neural networks
par: Yang, Yunfei, et autres
Publié: (2023)

Shallow diffusion networks provably learn hidden low-dimensional structure
par: Boffi, Nicholas M., et autres
Publié: (2024)

Unrolled denoising networks provably learn optimal Bayesian inference
par: Karan, Aayush, et autres
Publié: (2024)

MIQCQP reformulation of the ReLU neural networks Lipschitz constant estimation problem
par: Sbihi, Mohammed, et autres
Publié: (2024)

On bounds for norms of reparameterized ReLU artificial neural network parameters: sums of fractional powers of the Lipschitz norm control the network parameter vector
par: Jentzen, Arnulf, et autres
Publié: (2022)

On provable privacy vulnerabilities of graph representations
par: Wu, Ruofan, et autres
Publié: (2024)

Fixed-budget online adaptive learning for physics-informed neural networks. Towards parameterized problem inference
par: Nguyen, Thi Nguyen Khoa, et autres
Publié: (2022)

Towards provably efficient quantum algorithms for large-scale machine-learning models
par: Liu, Junyu, et autres
Publié: (2023)

Attention layers provably solve single-location regression
par: Marion, Pierre, et autres
Publié: (2024)

Generalization error bounds for two-layer neural networks with Lipschitz loss function
par: Nguwi, Jiang Yu, et autres
Publié: (2026)

A Lipschitz spaces view of infinitely wide shallow neural networks
par: Bartolucci, Francesca, et autres
Publié: (2024)

Learning efficient and provably convergent splitting methods
par: Kreusser, L. M., et autres
Publié: (2024)

Amortised and provably-robust simulation-based inference
par: Bharti, Ayush, et autres
Publié: (2026)

First-order ANIL provably learns representations despite overparametrization
par: Yüksel, Oğuz Kaan, et autres
Publié: (2023)

Ensembles provably learn equivariance through data augmentation
par: Nordenfors, Oskar, et autres
Publié: (2024)

Deep neural networks with ReLU, leaky ReLU, and softplus activation provably overcome the curse of dimensionality for Kolmogorov partial differential equations with Lipschitz nonlinearities in the $L^p$-sense
par: Ackermann, Julia, et autres
Publié: (2023)

An unsupervised tour through the hidden pathways of deep neural networks
par: Doimo, Diego
Publié: (2025)

Linear Recursive Feature Machines provably recover low-rank matrices
par: Radhakrishnan, Adityanarayanan, et autres
Publié: (2024)

Upper and lower bounds for the Lipschitz constant of random neural networks
par: Geuchen, Paul, et autres
Publié: (2023)

An efficient Wasserstein-distance approach for reconstructing jump-diffusion processes using parameterized neural networks
par: Xia, Mingtao, et autres
Publié: (2024)

Understanding polysemanticity in neural networks through coding theory
par: Marshall, Simon C., et autres
Publié: (2024)

TS-RSR: A provably efficient approach for batch Bayesian Optimization
par: Ren, Zhaolin, et autres
Publié: (2024)

Quantum automated learning with provable and explainable trainability
par: Ye, Qi, et autres
Publié: (2025)

Does provable absence of barren plateaus imply classical simulability?
par: Cerezo, M., et autres
Publié: (2023)

Dynamical simulation via quantum machine learning with provable generalization
par: Gibbs, Joe, et autres
Publié: (2022)

Reduced storage direct tensor ring decomposition for convolutional neural networks compression
par: Gabor, Mateusz, et autres
Publié: (2024)

Des-q: a quantum algorithm to provably speedup retraining of decision trees
par: Kumar, Niraj, et autres
Publié: (2023)

Explainability through uncertainty: Trustworthy decision-making with neural networks
par: Thuy, Arthur, et autres
Publié: (2024)

Resolution-free neural surrogates for geometric parameterization and mapping with spatially varying fields
par: Huang, Yanwen, et autres
Publié: (2026)

Equivariant score-based generative models provably learn distributions with symmetries efficiently
par: Chen, Ziyu, et autres
Publié: (2024)