Skip to content

prism-alignment

A dataset on language model preferences and value alignment

Visit official site
Category AI models

What prism-alignment is

The dataset is described as a collection for studying how large language models align with human preferences and values.

What you can do with it

  • Download a subset in a chosen format
  • Load and explore the data with pandas or mlcroissant

Getting started

  • Open the dataset page on Hugging Face
  • Download the data and filter variables for a research purpose

Pros & cons

✓ What we like

  • Loadable with pandas and mlcroissant
  • Released under a Creative Commons licence for research and education

! What to watch out for

  • The documented language data is English
  • Size is stated in a 10M to 100M range

FAQ

How was the data collected?

The material states surveys and multi-turn conversations gathered ratings and feedback.

What licence applies?

The material states a Creative Commons licence for research and educational use.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review