PPL
Yizhe Zhang
Papers (6)
DialoGPT: Large-Scale Generative Pre-training for Conversational Response GenerationLT2: Linear-Time Looped TransformersPrimal Generation, Dual Judgment: Self-Training from Test-Time ScalingSTARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal GenerationTrajectory as the Teacher: Few-Step Discrete Flow Matching via Energy-Navigated DistillationDACA-GRPO: Denoising-Aware Credit Assignment for Reinforcement Learning in Diffusion Language Models