Picture for Shengquan Chen

Shengquan Chen

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

Add code
May 20, 2024
Figure 1 for Lockpicking LLMs: A Logit-Based Jailbreak Using Token-level Manipulation
Figure 2 for Lockpicking LLMs: A Logit-Based Jailbreak Using Token-level Manipulation
Figure 3 for Lockpicking LLMs: A Logit-Based Jailbreak Using Token-level Manipulation
Figure 4 for Lockpicking LLMs: A Logit-Based Jailbreak Using Token-level Manipulation
Viaarxiv icon