Bottom-Up and Top-Down Analysis of Values, Agendas, and Observations in Corpora and LLMs
Computation and Language
2024-11-11 v1 Artificial Intelligence
Abstract
Large language models (LLMs) generate diverse, situated, persuasive texts from a plurality of potential perspectives, influenced heavily by their prompts and training data. As part of LLM adoption, we seek to characterize - and ideally, manage - the socio-cultural values that they express, for reasons of safety, accuracy, inclusion, and cultural fidelity. We present a validated approach to automatically (1) extracting heterogeneous latent value propositions from texts, (2) assessing resonance and conflict of values with texts, and (3) combining these operations to characterize the pluralistic value alignment of human-sourced and LLM-sourced textual data.
Keywords
Cite
@article{arxiv.2411.05040,
title = {Bottom-Up and Top-Down Analysis of Values, Agendas, and Observations in Corpora and LLMs},
author = {Scott E. Friedman and Noam Benkler and Drisana Mosaphir and Jeffrey Rye and Sonja M. Schmer-Galunder and Micah Goldwater and Matthew McLure and Ruta Wheelock and Jeremy Gottlieb and Robert P. Goldman and Christopher Miller},
journal= {arXiv preprint arXiv:2411.05040},
year = {2024}
}