paper-with-me

Papers

Mask in the Mirror: Implicit Sparsification

2024-08-19 · Tom Jacobs, Rebekka Burkholz

Sparsifying deep neural networks to reduce their inference cost is an NP-hard problem and difficult to optimize due to its mixed discrete and continuous nature. Yet, as we prove, continuous sparsification has already an implicit bias towards sparsity that would not require common projections of relaxed mask variables. While implicit rather than explicit regularization induces benefits, it usually does not provide enough flexibility in practice, as only a specific target sparsity is obtainable. To exploit its potential for continuous sparsification, we propose a way to control the strength of the implicit bias. Based on the mirror flow framework, we derive resulting convergence and optimality guarantees in the context of underdetermined linear regression and demonstrate the utility of our insights in more general neural network sparsification experiments, achieving significant performance gains, particularly in the high-sparsity regime. Our theoretical contribution might be of independent interest, as we highlight a way to enter the rich regime and show that implicit bias is controllable by a time-dependent Bregman potential.

📄 PDF Abstract BibTeX arXiv:2408.09966

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Linear Regression Linear Regression is a method for modelling a relationship between a dependent variable and independent variables. These models can be fit with numerous approaches. The most…

Similar Papers 제목 키워드 기반

Novel Gradient Sparsification Algorithm via Bayesian Inference

2024-09-23 · Ali Bereyhi, Ben Liang, Gary Boudreau, Ali Afana

Error accumulation is an essential component of the Top-$k$ sparsification method in distributed gradient descent. It implicitly scales the learning rate and prevents the slow-down of lateral movement, but it can also de…

Bayesian Inference

Effective Neural Network $L_0$ Regularization With BinMask

2023-04-21 · Kai Jia, Martin Rinard

$L_0$ regularization of neural networks is a fundamental problem. In addition to regularizing models for better generalizability, $L_0$ regularization also applies to selecting input features and training sparse neural n…

feature selection

Implicit Filter Sparsification In Convolutional Neural Networks

2019-05-13 · Dushyant Mehta, Kwang In Kim, Christian Theobalt

We show implicit filter level sparsity manifests in convolutional neural networks (CNNs) which employ Batch Normalization and ReLU activation, and are trained with adaptive gradient descent techniques and L2 regularizati…

L2 Regularization

Less Is More - On the Importance of Sparsification for Transformers and Graph Neural Networks for TSP

2024-03-25 · Attila Lischka, Jiaming Wu, Rafael Basso, Morteza Haghir Chehreghani 외

Most of the recent studies tackling routing problems like the Traveling Salesman Problem (TSP) with machine learning use a transformer or Graph Neural Network (GNN) based encoder architecture. However, many of them apply…

Graph Neural NetworkTraveling Salesman Problem

ONG: One-Shot NMF-based Gradient Masking for Efficient Model Sparsification

2025-08-18 · Sankar Behera, Yamuna Prasad arxiv

Deep Neural Networks (DNNs) have achieved remarkable success but their large size poses deployment challenges. While various pruning techniques exist, many involve complex iterative processes, specialized criteria, or st…