A unified data reconstruction attack achieves provable finite-width recovery in random feature networks and efficient subspace-based reconstruction for general models using weight changes.
Adversarial examples exist in two-layer relu networks for low dimensional linear subspaces.Advances in Neural Information Processing Systems, 36:5028–5049, 2023
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Efficient Techniques for Data Reconstruction, with Finite-Width Recovery Guarantees
A unified data reconstruction attack achieves provable finite-width recovery in random feature networks and efficient subspace-based reconstruction for general models using weight changes.