Cooperative Agency-Centered LLMs

Iyadunni J. Adenuga

Keywords: User Agency, AI Alignment

Abstract: AI models are not readily accepted and adopted in highly consequential fields like education, which cater to societal ideals. This is because injecting human values such as agency, which is an essential ingredient for learning, is hard. Using knowledge from cooperative AI and Liu et al. (2023) Stable Alignment method, we propose (without evaluation) a method for aligning large language models (LLMs) with the human agency of two different groups: teachers and students. This could ensure that effective learning occurs even with LLMs.

Submission Type: Tiny Paper (2 Pages)

Archival Option: This is an archival submission

Presentation Venue Preference: ICLR 2025

Submission Number: 88