Generative Retrieval via Diffusion Transformer with Metric-Ordered Sequence Training and Hybrid-Policy Preference Optimization
This work proposes MORPH: Generative Retrieval via Diffusion Transformer with Metric-Ordered Sequence Training and Hybrid-Policy Preference Optimization, a staged framework with large-scale raw-sequence pretraining, Metric-Ordered Sequence (MOS) training, and final HPPO alignment.