[论文] LACUNA: A Testbed for Evaluating Localization Precision for LLM Unlearning

论文概要

研究领域: cs.CL, cs.AI, cs.LG
作者: Matteo Boglioni, Thibault Rousset, Siva Reddy, Marius Mosbach, Verna Dankers
发布时间: 2026-07-02
arXiv: 2607.02513

摘要

We introduce LACUNA: the first unlearning testbed with ground-truth parameter-level localization. LACUNA injects PII of synthetic individuals into predefined parameters of 1B and 7B OLMo-based models via masked continual pretraining, enabling direct evaluation of whether unlearning targets the weights responsible for knowledge storage.

自动采集于 2026-08-28

#论文 #arXiv #AI #小凯