paper-with-me

Papers

Deconstructing Data Reconstruction: Multiclass, Weight Decay and General Losses

2023-09-21 · NeurIPS 2023 11

Memorization of training data is an active research area, yet our understanding of the inner workings of neural networks is still in its infancy. Recently, Haim et al. 2022 proposed a scheme to reconstruct training samples from multilayer perceptron binary classifiers, effectively demonstrating that a large portion of training samples are encoded in the parameters of such networks. In this work, we extend their findings in several directions, including reconstruction from multiclass and convolutional neural networks. We derive a more general reconstruction scheme which is applicable to a wider range of loss functions such as regression losses. Moreover, we study the various factors that contribute to networks' susceptibility to such reconstruction schemes. Intriguingly, we observe that using weight decay during training increases reconstructability both in terms of quantity and quality. Additionally, we examine the influence of the number of neurons relative to the number of training samples on the reconstructability. Code: https://github.com/gonbuzaglo/decoreco

📄 PDF Abstract BibTeX

Code (1)

gonbuzaglo/decoreco 공식 구현 pytorch

Methods 이 논문이 사용한 방법론

Weight Decay 설명 없음

Similar Papers 제목 키워드 기반

Reconstructing Training Data from Multiclass Neural Networks

2023-05-05 · Gon Buzaglo, Niv Haim, Gilad Yehudai, Gal Vardi 외

Reconstructing samples from the training set of trained neural networks is a major privacy concern. Haim et al. (2022) recently showed that it is possible to reconstruct training samples from neural network binary classi…

Binary Classification

Deep Learning with Eigenvalue Decay Regularizer

2016-04-24 · Oswaldo Ludwig

This paper extends our previous work on regularization of neural networks using Eigenvalue Decay by employing a soft approximation of the dominant eigenvalue in order to enable the calculation of its derivatives in relat…

ClassificationDeep LearningGeneral Classificationimage-classification+3

Identifying individual facial expressions by deconstructing a neural network

2016-06-23 · Farhad Arbabzadah, Grégoire Montavon, Klaus-Robert Müller, Wojciech Samek

This paper focuses on the problem of explaining predictions of psychological attributes such as attractiveness, happiness, confidence and intelligence from face photographs using deep neural networks. Since psychological…

AttributeGender PredictionTransfer Learning

The Geometry of Multi-Task Grokking: Transverse Instability, Superposition, and Weight Decay Phase Structure

2026-02-19 · Yongzhong Xu arxiv

Grokking -- the abrupt transition from memorization to generalization long after near-zero training loss -- has been studied mainly in single-task settings. We extend geometric analysis to multi-task modular arithmetic, …

Weighted MCC: A Robust Measure of Multiclass Classifier Performance for Observations with Individual Weights

2025-12-23 · Rommel Cortez, Bala Krishnamoorthy arxiv

Several performance measures are used to evaluate binary and multiclass classification tasks. But individual observations may often have distinct weights, and none of these measures are sensitive to such varying weights.…

Binary Classification