Foundational Challenges in Assuring Alignment and Safety of Large Language Models
It addresses critical safety and alignment problems for developers and users of large language models, but is incremental as it builds on existing concerns without new solutions.
The paper identifies 18 foundational challenges in ensuring the alignment and safety of large language models, categorizing them into scientific understanding, development/deployment methods, and sociotechnical aspects, and poses over 200 concrete research questions.
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.