Leveraging Per-Instance Privacy for Machine Unlearning @GoogleTechTalks
Leveraging Per-Instance Privacy for Machine Unlearning  @GoogleTechTalks
Uploaded January 2026 | Updated November 2026, 6 days ago
A Google TechTalk, 2025-09-03, presented by Anvith Thudi
Privacy in Machine Learning Seminar. ABSTRACT: We present a principled, per-instance approach to quantifying the difficulty of unlearning via finetuning. We begin by sharpening an analysis of noisy gradient descent for unlearning (Chien et al.,2024), obtaining a better utility–unlearning tradeoff by replacing worst-case privacy loss bounds with per-instance privacy losses (Thudi et al., 2024), each of which bounds the (Renyi) divergence to retraining without an individual data point. To demonstrate the practical applicability of our theory, we present empirical results showing that our theoretical predictions are born out both for Stochastic Gradient Langevin Dynamics (SGLD) as well as for standard fine-tuning without explicit noise. We further demonstrate that per-instance privacy losses correlate well with several existing data difficulty metrics, while also identifying harder groups of data points, and introduce novel evaluation methods based on loss barriers. All together, our findings provide a foundation for more efficient and adaptive unlearning strategies tailored to the unique properties of individual data points.
Leveraging Per-Instance Privacy for Machine Unlearning2023 Blockly Developer Summit Day 2-2: Blockly Themes for AccessibilityGoing Back and Beyond: Emerging (Old) Threats in LLM Privacy and PoisoningWhat Makes Software Work?Introduction to Length-Constrained Expanders and Expander DecompositionsOne Tree to Rule Them All: Polylogarithmic Universal Steiner TreesUnderstanding Trade-offs Between Worst-case Differential Privacy Guarantees & Real Threat ModelsGraph Attention RetrospectiveMapping from Training Data to Predictions with Datamodels2023 Blockly Developer Summit DAY 1-14: BlocksCAD - Math + Coding + DesignCharles Hoskinson | CEO of Input Output Global | web3 talks | Jan 5th 2023 | MC: Marlon RuizFixed-point Error Bounds for Mean-payoff Markov Decision Processes
Google TechTalks |

Leveraging Per-Instance Privacy for Machine Unlearning

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER