English
Related papers

Related papers: Lipschitz Constrained Parameter Initialization for…

200 papers

Scaling models has led to significant advancements in deep learning, but training these models in decentralized settings remains challenging due to communication bottlenecks. While existing compression techniques are effective in…

Machine Learning · Computer Science 2025-06-03 Sameera Ramasinghe , Thalaiyasingam Ajanthan , Gil Avraham , Yan Zuo , Alexander Long

It is notoriously difficult to train Transformers on small datasets; typically, large pre-trained models are instead used as the starting point. We explore the weights of such pre-trained Transformers (particularly for vision) to attempt to…

Computer Vision and Pattern Recognition · Computer Science 2023-05-18 Asher Trockman , J. Zico Kolter

Despite the recent success of stochastic gradient descent in deep learning, it is often difficult to train a deep neural network with an inappropriate choice of its initial parameters. Even if training is successful, it has been known that…

Machine Learning · Computer Science 2023-02-10 Cheolhyoung Lee , Kyunghyun Cho

Training a neural network (NN) depends on multiple factors, including but not limited to the initial weights. In this paper, we focus on initializing deep NN parameters such that it performs better, comparing to random or zero…

Machine Learning · Computer Science 2020-11-10 Mohamad H. Danesh

Understanding whether deep neural networks are effectively optimized remains challenging, as training occurs in highly nonconvex landscapes and standard metrics provide limited visibility into layer-wise learning quality. This challenge is…

Machine Learning · Computer Science 2026-05-05 Arian Eamaz , Farhang Yeganegi , Mojtaba Soltanalian

During pretraining, the Pre-LayerNorm transformer suffers from a gradient magnitude mismatch: gradients at early layers are much larger than at later layers. These issues can be alleviated by our proposed NormFormer architecture, which adds…

Computation and Language · Computer Science 2021-11-02 Sam Shleifer , Jason Weston , Myle Ott

Transformers process tokens in parallel but are temporally shallow: at position $t$, each layer attends to key-value pairs computed based on the previous layer, yielding a depth capped by the number of layers. Recurrent models offer…

Machine Learning · Computer Science 2026-04-24 Costin-Andrei Oncescu , Depen Morwani , Samy Jelassi , Alexandru Meterez , Mujin Kwun , Sham Kakade

Initialization of parameters in deep neural networks has been shown to have a big impact on the performance of the networks (Mishkin & Matas, 2015). The initialization scheme devised by He et al, allowed convolution activations to carry a…

Machine Learning · Computer Science 2017-02-28 Armen Aghajanyan

Neural networks are often highly sensitive to input and weight perturbations. This sensitivity has been linked to pathologies such as vulnerability to adversarial examples, divergent training, and overfitting. To combat these problems, past…

Machine Learning · Computer Science 2025-07-18 Laker Newhouse , R. Preston Hess , Franz Cesista , Andrii Zahorodnii , Jeremy Bernstein , Phillip Isola

Parameter fine tuning is a transfer learning approach whereby learned parameters from pre-trained source network are transferred to the target network followed by fine-tuning. Prior research has shown that this approach is capable of…

Computer Vision and Pattern Recognition · Computer Science 2019-09-20 Tasfia Shermin , Shyh Wei Teng , Manzur Murshed , Guojun Lu , Ferdous Sohel , Manoranjan Paul

Transformers have achieved great success in effectively processing sequential data such as text. Their architecture consisting of several attention and feedforward blocks can model relations between elements of a sequence in parallel…

Machine Learning · Computer Science 2025-02-20 Jaemu Heo , Eldor Fozilov , Hyunmin Song , Taehwan Kim

Deep neural networks achieve state-of-the-art performance for a range of classification and inference tasks. However, the use of stochastic gradient descent combined with the nonconvexity of the underlying optimization problems renders…

Machine Learning · Computer Science 2020-01-29 Ramina Ghods , Andrew S. Lan , Tom Goldstein , Christoph Studer

This paper discusses the weight parametrization of two standard 1-Lipschitz network architectures, the Almost-Orthogonal-Layers (AOL) and the SDP-based Lipschitz Layers (SLL). It examines their impact on initialization for deep 1-Lipschitz…

Machine Learning · Computer Science 2025-11-19 Marius F. R. Juston , Ramavarapu S. Sreenivas , William R. Norris , Dustin Nottage , Ahmet Soylemezoglu

Handling the ever-increasing scale of contemporary deep learning and transformer-based models poses a significant challenge. Overparameterized Transformer networks outperform prior art in Natural Language processing and Computer Vision.…

Machine Learning · Computer Science 2024-11-05 Soheil Zibakhsh Shabgahi , Mohammad Sohail Shariff , Farinaz Koushanfar

In recent years, transformer-based models have revolutionized deep learning, particularly in sequence modeling. To better understand this phenomenon, there is a growing interest in using Markov input processes to study transformers.…

Machine Learning · Computer Science 2024-06-28 Ashok Vardhan Makkuva , Marco Bondaschi , Chanakya Ekbote , Adway Girish , Alliot Nagle , Hyeji Kim , Michael Gastpar

One of the arguments to explain the success of deep learning is the powerful approximation capacity of deep neural networks. Such capacity is generally accompanied by the explosive growth of the number of parameters, which, in turn, leads…

Machine Learning · Computer Science 2022-09-15 Zuowei Shen , Haizhao Yang , Shijun Zhang

In the pursuit of reducing the number of trainable parameters in deep transformer networks, we employ Reinforcement Learning to dynamically select layers during training and tie them together. Every few iterations, the RL agent is asked…

Machine Learning · Computer Science 2024-01-24 Tamir David Hay , Lior Wolf

We present a new training methodology for transformers using a multilevel, layer-parallel approach. Through a neural ODE formulation of transformers, our application of a multilevel parallel-in-time algorithm for the forward and…

Machine Learning · Computer Science 2026-01-27 Shuai Jiang , Marc Salvadó-Benasco , Eric C. Cyr , Alena Kopaničáková , Rolf Krause , Jacob B. Schroder

LLM architecture research generally aims to maximize model quality subject to fixed compute/latency budgets. However, many applications of interest such as edge and on-device deployment are further constrained by the model's memory…

Machine Learning · Computer Science 2026-04-28 Abbas Zeitoun , Lucas Torroba-Hennigen , Yoon Kim

In this paper, we aim to build the global convergence theory of encoder-only shallow Transformers under a realistic setting from the perspective of architectures, initialization, and scaling under a finite width regime. The difficulty lies…

Machine Learning · Computer Science 2023-11-06 Yongtao Wu , Fanghui Liu , Grigorios G Chrysos , Volkan Cevher