English
Related papers

Related papers: Deconstructing the Goldilocks Zone of Neural Netwo…

200 papers

The proper initialization of weights is crucial for the effective training and fast convergence of deep neural networks (DNNs). Prior work in this area has mostly focused on balancing the variance among weights per layer to maintain…

Machine Learning · Computer Science 2020-06-05 Maciej Skorski , Alessandro Temperoni , Martin Theobald

We provide the first global optimization landscape analysis of $Neural\;Collapse$ -- an intriguing empirical phenomenon that arises in the last-layer classifiers and features of neural networks during the terminal phase of training. As…

Machine Learning · Computer Science 2021-05-07 Zhihui Zhu , Tianyu Ding , Jinxin Zhou , Xiao Li , Chong You , Jeremias Sulam , Qing Qu

Many aspects of the geometry of loss functions in deep learning remain mysterious. In this paper, we work toward a better understanding of the geometry of the loss function $L$ of overparameterized feedforward neural networks. In this…

Machine Learning · Computer Science 2020-05-19 Y. Cooper

The statistical properties of deep neural networks (DNNs) at initialization play an important role to comprehend their trainability and the intrinsic architectural biases they possess before data exposure Well established mean field (MF)…

Machine Learning · Computer Science 2026-03-03 Alberto Bassi , Marco Baity-Jesi , Aurelien Lucchi , Carlo Albert , Emanuele Francazi

The goal of this paper is to analyze the geometric properties of deep neural network classifiers in the input space. We specifically study the topology of classification regions created by deep networks, as well as their associated decision…

Computer Vision and Pattern Recognition · Computer Science 2017-05-29 Alhussein Fawzi , Seyed-Mohsen Moosavi-Dezfooli , Pascal Frossard , Stefano Soatto

Understanding the training dynamics of deep neural networks is challenging due to their high-dimensional nature and intricate loss landscapes. Recent studies have revealed that, along the training trajectory, the gradient approximately…

Machine Learning · Computer Science 2025-03-12 Minhak Song , Kwangjun Ahn , Chulhee Yun

Many of the recent remarkable advances in computer vision and language models can be attributed to the success of transfer learning via the pre-training of large foundation models. However, a theoretical framework which explains this…

Machine Learning · Computer Science 2024-12-19 Michael Munn , Benoit Dherin , Javier Gonzalvo

There are many surprising and perhaps counter-intuitive properties of optimization of deep neural networks. We propose and experimentally verify a unified phenomenological model of the loss landscape that incorporates many of them. High…

Machine Learning · Computer Science 2019-06-12 Stanislav Fort , Stanislaw Jastrzebski

We aim to understand grokking, a phenomenon where models generalize long after overfitting their training set. We present both a microscopic analysis anchored by an effective theory and a macroscopic analysis of phase diagrams describing…

Machine Learning · Computer Science 2022-10-17 Ziming Liu , Ouail Kitouni , Niklas Nolte , Eric J. Michaud , Max Tegmark , Mike Williams

The highly non-linear nature of deep neural networks causes them to be susceptible to adversarial examples and have unstable gradients which hinders interpretability. However, existing methods to solve these issues, such as adversarial…

Machine Learning · Computer Science 2023-01-11 Suraj Srinivas , Kyle Matoba , Himabindu Lakkaraju , Francois Fleuret

Robustness of deep neural networks to input noise remains a critical challenge, as naive noise injection often degrades accuracy on clean (uncorrupted) data. We propose a novel training framework that addresses this trade-off through two…

Machine Learning · Statistics 2026-01-06 Hai-Vy Nguyen , Fabrice Gamboa , Sixin Zhang , Reda Chhaibi , Serge Gratton , Thierry Giaccone

The training dynamics of deep neural networks often defy expectations, even as these models form the foundation of modern machine learning. Two prominent examples are grokking, where test performance improves abruptly long after the…

Machine Learning · Computer Science 2026-01-28 Keitaro Sakamoto , Issei Sato

We introduce the new "Goldilocks" class of activation functions, which non-linearly deform the input signal only locally when the input signal is in the appropriate range. The small local deformation of the signal enables better…

Machine Learning · Computer Science 2021-10-11 Jan Rosenzweig , Zoran Cvetkovic , Ivana Rosenzweig

Prior work has demonstrated a consistent tendency in neural networks engaged in continual learning tasks, wherein intermediate task similarity results in the highest levels of catastrophic interference. This phenomenon is attributed to the…

Uncertainty calibration is crucial for various machine learning applications, yet it remains challenging. Many models exhibit hallucinations - confident yet inaccurate responses - due to miscalibrated confidence. Here, we show that the…

Machine Learning · Computer Science 2025-03-28 Jeonghwan Cheon , Se-Bum Paik

Training neural networks with first order optimisation methods is at the core of the empirical success of deep learning. The scale of initialisation is a crucial factor, as small initialisations are generally associated to a feature…

Machine Learning · Computer Science 2025-09-16 Etienne Boursier , Nicolas Flammarion

We revisit and extend the experiments of Goodfellow et al. (2014), who showed that - for then state-of-the-art networks - "the objective function has a simple, approximately convex shape" along the linear path between initialization and the…

Machine Learning · Computer Science 2020-12-15 Jonathan Frankle

Plasticity, the ability of a neural network to quickly change its predictions in response to new information, is essential for the adaptability and robustness of deep reinforcement learning systems. Deep neural networks are known to lose…

Machine Learning · Computer Science 2023-11-28 Clare Lyle , Zeyu Zheng , Evgenii Nikishin , Bernardo Avila Pires , Razvan Pascanu , Will Dabney

Neural networks trained via gradient descent with random initialization and without any regularization enjoy good generalization performance in practice despite being highly overparametrized. A promising direction to explain this phenomenon…

Machine Learning · Computer Science 2022-05-17 Hancheng Min , Salma Tarmoun , Rene Vidal , Enrique Mallada

Generalization of deep neural networks remains one of the main open problems in machine learning. Previous theoretical works focused on deriving tight bounds of model complexity, while empirical works revealed that neural networks exhibit…

Machine Learning · Computer Science 2022-01-31 James Wang , Cheng-Lin Yang