Salvato in:
Dettagli Bibliografici
Autori principali: Bell, Henry, Schertel, Lara Neubauer da Costa, Ding, Bochu, Fain, Brandon
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2601.18760
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918306038939648
author Bell, Henry
Schertel, Lara Neubauer da Costa
Ding, Bochu
Fain, Brandon
author_facet Bell, Henry
Schertel, Lara Neubauer da Costa
Ding, Bochu
Fain, Brandon
contents A crucial consideration when developing and deploying Large Language Models (LLMs) is the human values to which these models are aligned. In the constitutional framework of alignment models are aligned to a set of principles (the constitution) specified in natural language. However, it is unclear how to fairly determine this constitution with widespread stakeholder input. In this work we propose Grounded Constitutional AI (GCAI), a unified framework for generating constitutions of principles that are representative of both users' general expectations toward AI (general principles) and their interaction-time preferences (contextual principles). We extend the Inverse Constitutional AI (ICAI) approach to generate contextual principles from human preference annotation data by leveraging human-provided \textit{reasons} for their preferences. We supplement these contextual principles with general principles surfaced from user statements of \textit{values} regarding AI. We show that a constitution generated by GCAI is preferred by humans over one generated through ICAI both personally, and for widespread use in governing AI behavior. Additionally participants consider the GCAI constitution to be more morally grounded, coherent, and pluralistic.
format Preprint
id arxiv_https___arxiv_org_abs_2601_18760
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Beyond Preferences: Learning Alignment Principles Grounded in Human Reasons and Values
Bell, Henry
Schertel, Lara Neubauer da Costa
Ding, Bochu
Fain, Brandon
Machine Learning
Computation and Language
A crucial consideration when developing and deploying Large Language Models (LLMs) is the human values to which these models are aligned. In the constitutional framework of alignment models are aligned to a set of principles (the constitution) specified in natural language. However, it is unclear how to fairly determine this constitution with widespread stakeholder input. In this work we propose Grounded Constitutional AI (GCAI), a unified framework for generating constitutions of principles that are representative of both users' general expectations toward AI (general principles) and their interaction-time preferences (contextual principles). We extend the Inverse Constitutional AI (ICAI) approach to generate contextual principles from human preference annotation data by leveraging human-provided \textit{reasons} for their preferences. We supplement these contextual principles with general principles surfaced from user statements of \textit{values} regarding AI. We show that a constitution generated by GCAI is preferred by humans over one generated through ICAI both personally, and for widespread use in governing AI behavior. Additionally participants consider the GCAI constitution to be more morally grounded, coherent, and pluralistic.
title Beyond Preferences: Learning Alignment Principles Grounded in Human Reasons and Values
topic Machine Learning
Computation and Language
url https://arxiv.org/abs/2601.18760