Arithmetic Control of LLMs for Diverse User Preferences: Directional Preference Alignment with Multi-Objective Rewards

arXiv:2402.18571v1 Announce Type: new
Abstract: Fine-grained control over large language models (LLMs) remains a significant challenge, hindering their adaptability to diverse user needs. While Reinforcement Learning from Human Feedback (RLHF) shows promise in aligning LLMs, its reliance on scalar rewards often limits its ability to capture diverse user preferences in real-world applications. To address this limitation, we introduce the Directional Preference Alignment (DPA) framework. Unlike the scalar-reward RLHF, DPA incorporates multi-objective reward modeling to represent diverse preference profiles. Additionally, DPA …

abstract adaptability alignment arxiv challenge control cs.ai cs.cl cs.lg diverse feedback fine-grained human human feedback language language models large language large language models llms multi-objective reinforcement reinforcement learning reliance rlhf shows stat.ml type