Research map / Federated learning and privacy / Differential privacy
Remember What You Want to Forget: Algorithms for Machine Unlearning
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 |
|---|---|---|---|
| Use the requested deleted samples in a Newton-style update instead of having the unlearning procedure ignore them | Machine unlearning for convex losses · Convex losses | deletion capacity: O(n/d^{1/4}) vs. \Theta(n/d^{1/2}) | Helped |
| Remove samples by exactly updating the empirical minimizer | Mean estimation with squared loss · Bernoulli(1/2) mean estimation with adversarial deletions | test loss: m/n | Hurt |
| Approximate the minimizer after deletion with a Newton update, then add Gaussian noise scaled to the approximation error | Newton-update unlearning · Strongly convex losses | deletion capacity: O(m^2 / n^2) | Helped |
| Use the empirical-risk minimizer after deletion as the unlearning output | Mean estimation with squared loss · Empirical loss on the remaining training samples | empirical loss | Helped |
Closest papers by meaning
- The Utility and Complexity of In- and Out-of-Distribution Machine Unlearning (ICLR 2025 · similarity 0.93)
- Rewind-to-Delete: Certified Machine Unlearning for Nonconvex Functions (NeurIPS 2025 · similarity 0.92)
- Towards Source-Free Machine Unlearning (CVPR 2025 · similarity 0.91)
- Langevin Unlearning: A New Perspective of Noisy Gradient Descent for Machine Unlearning (NeurIPS 2024 · similarity 0.91)
- Leveraging Per-Instance Privacy for Machine Unlearning (ICML 2025 · similarity 0.91)
- Certified Unlearning for Neural Networks (ICML 2025 · similarity 0.90)
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.