Reproduce “A Scale For Value Alignment In Human-AI Interaction” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Lena Hegemann, Steeven Villa, Hyemin Bang, Mitchell L. Gordon, Antti Oulasvirta, Robin Welsch, Patrick Ebel
The Perceived Value Alignment Scale measures whether users think an AI agent understands their values and acts in accordance with them. It separates Value Understanding from Value Manifestation in a 13-item questionnaire. In tests with 658 participants, it distinguished higher- and lower-alignment versions of car and travel scenarios, though it measures perceptions after scenario exposure, not objective or deployed alignment.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.