ICML26: Asymmetric Sources Unlearning Improves Privacy-Utility Tradeoff
gkdziugaite · x · 2026-07-07
At ICML 2026 in Seoul, researchers presented the paper "Unlearning with Asymmetric Sources," proposing the use of public data to improve the tradeoff between privacy certification and model utility in Machine Unlearning. Machine unlearning aims to make models forget specific training data and is a key technology for AI privacy compliance. This study provides a new theoretical framework for large-scale unlearning scenarios.
Related event: ICML 2026 Paper Proposes ALU Framework for Large-Scale Machine Unlearning(5 posts)→
More from Safety
- DHH Slams 'GDPR Is Good' Take: Vague Rules Birthed a Bureaucratic Beast — dhh · 2026-09-11
- Houthis tried to use Claude to design missile software, Anthropic says it blocked the attempts — Affectionate_Bee6434 · 2026-09-11
- AI safety community mocked as 'bridge engineers' who say bridges can never be safe — Dan_Jeffries1 · 2026-09-11
- Why So Many AI Researchers Think the Machines Could Kill Everyone — wiredmagazine · 2026-09-11
- California creates standards for independent AI auditors to verify lab safety testing — VraserX · 2026-09-11
- a16z podcast: why 2-3 person startups are absent from policy debates — a16z Podcast · 2026-09-11