"Lockpicking LLMs: A Logit-Based Jailbreak Using Token-level Manipulation."

Yuxi Li et al. (2024)

Details and statistics

DOI: 10.48550/ARXIV.2405.13068

access: open

type: Informal or Other Publication

metadata version: 2024-08-08

a service of  Schloss Dagstuhl - Leibniz Center for Informatics