The PRISM alignment dataset: What participatory, representative and individualised human feedback reveals about the subjective and multicultural alignment of large ...
Authors: HR Kirk, M Bartolo, A Whitefield, P Rottger
Published: 2024
Publication: Advances in ..., 2024 - proceedings.neurips.cc
The PRISM Alignment Dataset presents a large-scale, culturally diverse human feedback dataset linking sociodemographic profiles of 1,500 participants from 75 countries to their contextual preferences and fine‑grained ratings in 8,011 live conversations with 21 LLMs. This enables analysis of how subjective values vary across people and cultures in LLM alignment data.
Institution: Meta, Cohere, AWS AI Labs, Contextual AI, Factored AI, University of Oxford, Bocconi University, Meedan, Hugging Face, University College London, ML Commons, University of Pennsylvania
Research Area: LLM Alignment, Human Feedback,Multicultural Studies
Discipline: Artificial Intelligence , Computational Social Science
Citations: 204
DOI: https://doi.org/10.52202/079017-3342