English
Related papers

Related papers: Overparametrization bends the landscape: BBP trans…

200 papers

The clear understanding of the non-convex landscape of neural network is a complex incomplete problem. This paper studies the landscape of linear (residual) network, the simplified version of the nonlinear network. By treating the gradient…

Algebraic Geometry · Mathematics 2021-02-09 Xiuyi Yang

Barren plateaus (BPs) are usually described by the exponential suppression of gradient variance, but the mechanism by which gradient signal disappears remains unclear. We show that this phenomenon can be understood as destructive…

Quantum Physics · Physics 2026-05-05 Pilsung Kang

Batch Normalization (BatchNorm) is an extremely useful component of modern neural network architectures, enabling optimization using higher learning rates and achieving faster convergence. In this paper, we use mean-field theory to…

Machine Learning · Computer Science 2019-03-08 Mingwei Wei , James Stokes , David J Schwab

A recent line of work studies overparametrized neural networks in the "kernel regime," i.e. when the network behaves during training as a kernelized linear predictor, and thus training with gradient descent has the effect of finding the…

Machine Learning · Computer Science 2020-02-26 Blake Woodworth , Suriya Gunasekar , Pedro Savarese , Edward Moroshko , Itay Golan , Jason Lee , Daniel Soudry , Nathan Srebro

Variational Bayesian posterior inference often requires simplifying approximations such as mean-field parametrisation to ensure tractability. However, prior work has associated the variational mean-field approximation for Bayesian neural…

Machine Learning · Computer Science 2022-10-07 Richard Kurle , Ralf Herbrich , Tim Januschowski , Yuyang Wang , Jan Gasthaus

One of the most surprising and exciting discoveries in supervised learning was the benefit of overparameterization (i.e. training a very large model) to improving the optimization landscape of a problem, with minimal effect on statistical…

Machine Learning · Statistics 2020-07-17 Rares-Darius Buhai , Yoni Halpern , Yoon Kim , Andrej Risteski , David Sontag

We study the properties of common loss surfaces through their Hessian matrix. In particular, in the context of deep learning, we empirically show that the spectrum of the Hessian is composed of two parts: (1) the bulk centered near zero,…

Machine Learning · Computer Science 2018-05-08 Levent Sagun , Utku Evci , V. Ugur Guney , Yann Dauphin , Leon Bottou

Modern deep learning models generalize remarkably well in-distribution, despite being overparametrized and trained with little to no explicit regularization. Instead, current theory credits implicit regularization imposed by the choice of…

Machine Learning · Computer Science 2026-03-17 Jonathan Wenger , Beau Coker , Juraj Marusic , John P. Cunningham

Many modern learning tasks involve fitting nonlinear models to data which are trained in an overparameterized regime where the parameters of the model exceed the size of the training dataset. Due to this overparameterization, the training…

Machine Learning · Computer Science 2018-12-27 Samet Oymak , Mahdi Soltanolkotabi

Deep Metric Learning trains a neural network to map input images to a lower-dimensional embedding space such that similar images are closer together than dissimilar images. When used for item retrieval, a query image is embedded using the…

Computer Vision and Pattern Recognition · Computer Science 2022-10-05 Konstantin Kobs , Andreas Hotho

Hyperbolic neural networks (HNNs) have demonstrated notable efficacy in representing real-world data with hierarchical structures via exploiting the geometric properties of hyperbolic spaces characterized by negative curvatures. Curvature…

Machine Learning · Computer Science 2025-08-27 Xiaomeng Fan , Yuwei Wu , Zhi Gao , Mehrtash Harandi , Yunde Jia

Despite its long history, Bayesian neural networks (BNNs) and variational training remain underused in practice: standard Gaussian posteriors misalign with network geometry, KL terms can be brittle in high dimensions, and implementations…

Machine Learning · Computer Science 2025-09-09 Carlos Stein Brito

Understanding the structure of neural network loss surfaces, particularly the emergence of low-loss tunnels, is critical for advancing neural network theory and practice. In this paper, we propose a novel approach to directly embed loss…

Machine Learning · Computer Science 2025-03-06 Daniel Dold , Julius Kobialka , Nicolai Palm , Emanuel Sommer , David Rügamer , Oliver Dürr

We study how the optimization landscape of a neural network deforms as a non-linear activation is introduced through a smooth homotopy. Working first in an abstract local setting - a smooth one-parameter family of objective functions…

Optimization and Control · Mathematics 2026-04-03 Michael Alexander

We study randomly initialized residual networks using mean field theory and the theory of difference equations. Classical feedforward neural networks, such as those with tanh activations, exhibit exponential behavior on the average when…

Neural and Evolutionary Computing · Computer Science 2017-12-27 Greg Yang , Samuel S. Schoenholz

Bayesian Neural Networks with Latent Variables (BNN+LVs) capture predictive uncertainty by explicitly modeling model uncertainty (via priors on network weights) and environmental stochasticity (via a latent input noise variable). In this…

Machine Learning · Computer Science 2022-09-19 Yaniv Yacoby , Weiwei Pan , Finale Doshi-Velez

The loss landscape of neural networks is a critical aspect of their training, and understanding its properties is essential for improving their performance. In this paper, we investigate how the loss surface changes when the sample size…

Machine Learning · Computer Science 2024-09-19 Nikita Kiselev , Andrey Grabovoy

In supervised learning, it is known that overparameterized neural networks with one hidden layer provably and efficiently learn and generalize, when trained using stochastic gradient descent with a sufficiently small learning rate and…

Machine Learning · Computer Science 2022-03-24 Kulin Shah , Amit Deshpande , Navin Goyal

Blind deconvolution involves the estimation of a sharp signal or image given only a blurry observation. Because this problem is fundamentally ill-posed, strong priors on both the sharp image and blur kernel are required to regularize the…

Computer Vision and Pattern Recognition · Computer Science 2013-05-13 David Wipf , Haichao Zhang

We discuss several aspects of the loss landscape of regularized neural networks: the structure of stationary points, connectivity of optimal solutions, path with nonincreasing loss to arbitrary global optimum, and the nonuniqueness of…

Machine Learning · Computer Science 2025-04-30 Sungyoon Kim , Aaron Mishkin , Mert Pilanci