"Improving Activation Steering in Language Models with Mean-Centring."

Ole Jorgensen et al. (2023)

Details and statistics

DOI: 10.48550/ARXIV.2312.03813

access: open

type: Informal or Other Publication

metadata version: 2024-05-24

a service of  Schloss Dagstuhl - Leibniz Center for Informatics