ValueCompass: A Framework for Measuring Contextual Value Alignment Between Human and LLMs

TL;DR

ValueCompass is a framework grounded in psychological theory that systematically measures how well large language models align with fundamental human values across different real-world scenarios, revealing notable misalignments.

Contribution

It introduces a novel, theory-based framework for assessing human-AI value alignment and applies it to LLMs in diverse contexts, highlighting the importance of context-aware alignment strategies.

Findings

Humans often endorse values like 'National Security' that LLMs reject.

Values differ significantly across scenarios, affecting alignment.

The framework reveals substantial misalignments needing address.

Abstract

As AI systems become more advanced, ensuring their alignment with a diverse range of individuals and societal values becomes increasingly critical. But how can we capture fundamental human values and assess the degree to which AI systems align with them? We introduce ValueCompass, a framework of fundamental values, grounded in psychological theory and a systematic review, to identify and evaluate human-AI alignment. We apply ValueCompass to measure the value alignment of humans and large language models (LLMs) across four real-world scenarios: collaborative writing, education, public sectors, and healthcare. Our findings reveal concerning misalignments between humans and LLMs, such as humans frequently endorse values like "National Security" which were largely rejected by LLMs. We also observe that values differ across scenarios, highlighting the need for context-aware AI alignment strategies. This work provides valuable insights into the design space of human-AI alignment, laying the foundations for developing AI systems that responsibly reflect societal values and ethics.