SAR

A Framework for Socially Assistive Tasks

Design considerations for developing robots to help people.

Reinforcement Learning from Human Feedback

How reinforcement learning with human feedback can help on the alignment problem.