3 papers
cs.LG2024
What makes unlearning hard and what to do about it
Kairan Zhao, Meghdad Kurmanji, George-Octavian BÄrbulescu +2
Machine unlearning is the problem of removing the effect of a subset of training data (the ''forget set'') from a trained model without damaging the model's utility e.g. to comply…
cs.DB2024
Learned Graph Rewriting with Equality Saturation: A New Paradigm in Relational Query Rewrite and Beyond
George-Octavian BÄrbulescu, Taiyi Wang, Zak Singh +1
Query rewrite systems perform graph substitutions using rewrite rules to generate optimal SQL query plans. Rewriting logical and physical relational query plans is proven to be an…
cs.LG2024
To Each (Textual Sequence) Its Own: Improving Memorized-Data Unlearning in Large Language Models
George-Octavian Barbulescu, Peter Triantafillou
LLMs have been found to memorize training textual sequences and regurgitate verbatim said sequences during text generation time. This fact is known to be the cause of privacy and r…