TroubleLLM: Align to Red Team Expert

arXiv:2403.00829v1 Announce Type: cross
Abstract: Large Language Models (LLMs) become the start-of-the-art solutions for a variety of natural language tasks and are integrated into real-world applications. However, LLMs can be potentially harmful in manifesting undesirable safety issues like social biases and toxic content. It is imperative to assess its safety issues before deployment. However, the quality and diversity of test prompts generated by existing methods are still far from satisfactory. Not only are these methods labor-intensive and require large budget …

abstract applications art arxiv become biases cs.ai cs.cl deployment expert language language models large language large language models llms natural natural language red team safety social solutions tasks team type world