Foundational Challenges in Assuring Alignment and Safety of Large Language Models

arXiv:2404.09932v1 Announce Type: new
Abstract: This work identifies 18 foundational challenges in assuring the alignment and safety of large language models (LLMs). These challenges are organized into three different categories: scientific understanding of LLMs, development and deployment methods, and sociotechnical challenges. Based on the identified challenges, we pose $200+$ concrete research questions.

abstract alignment arxiv challenges cs.ai cs.cl cs.cy cs.lg deployment development foundational language language models large language large language models llms safety scientific type understanding work