English
Related papers

Related papers: Training Dynamics of Multi-Head Softmax Attention …

200 papers

Flow matching models have recently emerged as an efficient alternative to diffusion, especially for text-guided image generation and editing, offering faster inference through continuous-time dynamics. However, existing flow-based editors…

Computer Vision and Pattern Recognition · Computer Science 2026-02-11 Carmine Zaccagnino , Fabio Quattrini , Enis Simsar , Marta Tintoré Gazulla , Rita Cucchiara , Alessio Tonioni , Silvia Cascianelli

Transformer models exhibit in-context learning: the ability to accurately predict the response to a novel query based on illustrative examples in the input sequence. In-context learning contrasts with traditional in-weights learning of…

Machine Learning · Computer Science 2023-12-07 Gautam Reddy

This paper formulates a multitask optimization problem where agents in the network have individual objectives to meet, or individual parameter vectors to estimate, subject to a smoothness condition over the graph. The smoothness condition…

Multiagent Systems · Computer Science 2019-11-18 Roula Nassif , Stefan Vlaski , Cedric Richard , Ali H. Sayed

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

Softmax is widely used in neural networks for multiclass classification, gate structure and attention mechanisms. The statistical assumption that the input is normal distributed supports the gradient stability of Softmax. However, when used…

Computer Vision and Pattern Recognition · Computer Science 2021-08-17 Shulun Wang , Bin Liu , Feng Liu

Multitask learning is a methodology to boost generalization performance and also reduce computational intensity and memory usage. However, learning multiple tasks simultaneously can be more difficult than learning a single task because it…

Machine Learning · Computer Science 2020-06-03 Sungjae Lee , Youngdoo Son

By jointly learning multiple tasks, multi-task learning (MTL) can leverage the shared knowledge across tasks, resulting in improved data efficiency and generalization performance. However, a major challenge in MTL lies in the presence of…

Machine Learning · Computer Science 2024-07-03 Hao Ban , Kaiyi Ji

The intriguing in-context learning (ICL) abilities of deep Transformer models have lately garnered significant attention. By studying in-context linear regression on unimodal Gaussian data, recent empirical and theoretical works have argued…

Machine Learning · Computer Science 2024-10-30 Khashayar Gatmiry , Nikunj Saunshi , Sashank J. Reddi , Stefanie Jegelka , Sanjiv Kumar

We study the fundamental optimization principles of self-attention, the defining mechanism of transformers, by analyzing the implicit bias of gradient-based optimizers in training a self-attention layer with a linear decoder in binary…

Machine Learning · Computer Science 2025-04-01 Bhavya Vasudeva , Puneesh Deora , Christos Thrampoulidis

We perform a convergence analysis of a discrete-in-time minimization scheme approximating a finite dimensional singularly perturbed gradient flow. We allow for different scalings between the viscosity parameter $\varepsilon$ and the time…

Analysis of PDEs · Mathematics 2018-11-14 Giovanni Scilla , Francesco Solombrino

We study the convergence properties of gradient descent for training deep linear neural networks, i.e., deep matrix factorizations, by extending a previous analysis for the related gradient flow. We show that under suitable conditions on…

Machine Learning · Computer Science 2021-11-25 Gabin Maxime Nguegnang , Holger Rauhut , Ulrich Terstiege

We study the \emph{in-context learning} (ICL) ability of a \emph{Linear Transformer Block} (LTB) that combines a linear attention component and a linear multi-layer perceptron (MLP) component. For ICL of linear regression with a Gaussian…

Machine Learning · Statistics 2024-02-26 Ruiqi Zhang , Jingfeng Wu , Peter L. Bartlett

Hyperparameter tuning is one of the essential steps to guarantee the convergence of machine learning models. We argue that intuition about the optimal choice of hyperparameters for stochastic gradient descent can be obtained by studying a…

Disordered Systems and Neural Networks · Physics 2025-12-12 Chanju Park , Biagio Lucini , Gert Aarts

Multi-task learning shows strikingly inconsistent results -- sometimes joint training helps substantially, sometimes it actively harms performance -- yet the field lacks a principled framework for predicting these outcomes. We identify a…

Machine Learning · Computer Science 2026-04-10 Jasper Zhang , Bryan Cheng

The training process of ReLU neural networks often exhibits complicated nonlinear phenomena. The nonlinearity of models and non-convexity of loss pose significant challenges for theoretical analysis. Therefore, most previous theoretical…

Machine Learning · Computer Science 2023-12-29 Mingze Wang , Chao Ma

This paper presents a novel optimization method for maximizing generalization over tasks in meta-learning. The goal of meta-learning is to learn a model for an agent adapting rapidly when presented with previously unseen tasks. Tasks are…

Machine Learning · Computer Science 2018-10-19 Amir Erfan Eshratifar , David Eigen , Massoud Pedram

We investigate the mechanistic underpinnings of in-context learning (ICL) in large language models by reconciling two dominant perspectives: the component-level analysis of attention heads and the holistic decomposition of ICL into Task…

Computation and Language · Computer Science 2026-05-04 Haolin Yang , Hakaze Cho , Naoya Inoue

In both Computer Vision and the wider Deep Learning field, the Transformer architecture is well-established as state-of-the-art for many applications. For Multitask Learning, however, where there may be many more queries necessary compared…

Computer Vision and Pattern Recognition · Computer Science 2025-08-07 Christian Bohn , Thomas Kurbiel , Klaus Friedrichs , Hasan Tercan , Tobias Meisen

The training of neural networks by gradient descent methods is a cornerstone of the deep learning revolution. Yet, despite some recent progress, a complete theory explaining its success is still missing. This article presents, for…

Machine Learning · Statistics 2026-04-15 Etienne Boursier , Loucas Pillaud-Vivien , Nicolas Flammarion

Softmax attention is a central component of transformer architectures, yet its nonlinear structure poses significant challenges for theoretical analysis. We develop a unified, measure-based framework for studying single-layer softmax…

Machine Learning · Computer Science 2025-12-15 Etienne Boursier , Claire Boyer