Simulating Training Dynamics to Reconstruct Training Data from Deep Neural Networks
In this work, we propose SimuDy to reconstruct training data from the parameters of DNNs
Apr 25, 2025
Unified Gradient-Based Machine Unlearning with Remain Geometry Enhancement
This work proposes a fast-slow parameter update strategy to implicitly approximate the up-to-date salient unlearning direction, free from specific modal constraints, and adaptable across computer vision unlearning tasks, including classification and generation.
Oct 7, 2024
Trainable Weight Averaging: Efficient Training by Optimizing Historical Solutions
A parallel framework for large-scale training with efficiency in memory and computation is designed for TWA or EMA and manifests better adaptation to different stages of training.
Feb 26, 2023
Adversarial Attack on Attackers: Post-Process to Mitigate Black-Box Score-Based Query Attacks
We propose a novel defense against score-based query attacks, which post-processes model outputs to effectively confound attackers without hurting accuracy and calibration.
Nov 1, 2022