4 papers
Understanding the Performance Gap in Preference Learning: A Dichotomy of RLHF and DPO
Ruizhe Shi, Minhak Song, Runlong Zhou +3
We present a fine-grained theoretical analysis of the performance gap between two-stage reinforcement learning from human feedback~(RLHF) and direct preference optimization~(DPO).…
The Crucial Role of Samplers in Online Direct Preference Optimization
Ruizhe Shi, Runlong Zhou, Simon S. Du
Direct Preference Optimization (DPO) has emerged as a stable, scalable, and efficient solution for language model alignment. Despite its empirical success, the optimization propert…
Decoding-Time Language Model Alignment with Multiple Objectives
Ruizhe Shi, Yifang Chen, Yushi Hu +4
Aligning language models (LMs) to human preferences has emerged as a critical pursuit, enabling these models to better serve diverse user needs. Existing methods primarily focus on…
Rethinking Transformers in Solving POMDPs
Chenhao Lu, Ruizhe Shi, Yuyao Liu +3
Sequential decision-making algorithms such as reinforcement learning (RL) in real-world scenarios inevitably face environments with partial observability. This paper scrutinizes th…