Grok-2 Beta The training set data can come from x.AI (X.AI, 2024) which was formerly Twitter

Limitation As mentioned before, because the experimental tools GPT4o, Grok-2 Beta, Llama 3 · 2019

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

browse 1 citing papers

representative citing papers

cs.CR · 2024-11-23 · conditional · novelty 3.0

All five tested LLMs generated verbal abuse after a multi-step moralizing jailbreak; Claude 3.5 Sonnet appeared most resistant.

Showing 1 of 1 citing paper.

"Moralized" Multi-Step Jailbreak Prompts: Black-Box Testing of Guardrails in Large Language Models for Verbal Attacks cs.CR · 2024-11-23 · conditional · none · ref 5
All five tested LLMs generated verbal abuse after a multi-step moralizing jailbreak; Claude 3.5 Sonnet appeared most resistant.