Adaptive Alignment: Dynamic Preference Adjustments via Multi-Objective Reinforcement Learning for Pluralistic AI

Autor:	Harland, Hadassah, Dazeley, Richard, Vamplew, Peter, Senaratne, Hashini, Nakisa, Bahareh, Cruz, Francisco
Rok vydání:	2024
Předmět:	Computer Science - Machine Learning Computer Science - Artificial Intelligence
Druh dokumentu:	Working Paper
Popis:	Emerging research in Pluralistic Artificial Intelligence (AI) alignment seeks to address how intelligent systems can be designed and deployed in accordance with diverse human needs and values. We contribute to this pursuit with a dynamic approach for aligning AI with diverse and shifting user preferences through Multi Objective Reinforcement Learning (MORL), via post-learning policy selection adjustment. In this paper, we introduce the proposed framework for this approach, outline its anticipated advantages and assumptions, and discuss technical details about the implementation. We also examine the broader implications of adopting a retroactive alignment approach through the sociotechnical systems perspective. Comment: Accepted for the Pluralistic Alignment workshop at NeurIPS 2024
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2410.23630 Zobrazit plný text záznamu View this record from Arxiv