Research map / Label-efficient and robust learning / Recognition and long-tail
Reconstructing Training Data From Trained Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| reconstruct via KKT stationarity loss instead of model inversion | Reconstruction scheme · 2D toy circle | training samples reconstructed: 7 -> 20 | Helped |
| reconstruct via KKT stationarity loss instead of model inversion | Reconstruction scheme · MNIST/CIFAR10 | reconstruction quality | Helped |
| use small initialization of first-layer weights (std 1e-4) (ablation) | trained MLP · MNIST/CIFAR10 | reconstruction quality | Helped |
| train for more epochs (ablation) | trained MLP · MNIST/CIFAR10 | reconstruction quality | Helped |
| increase number of training samples n (ablation) | trained MLP · CIFAR10 | reconstruction SSIM | Hurt |
| use non-homogeneous MLP with biases in all hidden layers (ablation) | trained MLP · CIFAR10 | reconstruction SSIM | No clear effect |
Closest papers by meaning
- Understanding Reconstruction Attacks with the Neural Tangent Kernel and Dataset Distillation (ICLR 2024 · similarity 0.85)
- Certified Unlearning for Neural Networks (ICML 2025 · similarity 0.83)
- Network Inversion of Convolutional Neural Nets (Student Abstract) (AAAI 2025 · similarity 0.83)
- High-Fidelity Gradient Inversion in Distributed Learning (AAAI 2024 · similarity 0.83)
- A Decade's Battle on Dataset Bias: Are We There Yet? (ICLR 2025 · similarity 0.83)
- Trained Random Forests Completely Reveal your Dataset (ICML 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.