Sanqiang Zhao

text generation (4) text simplification (2) data augmentation (2) transfer learning (1) lexical complexity (1) dialogue generation (1) knowledge distillation (1) preference optimization (1) multimodal learning (1) image captioning (1) reinforcement learning from human feedback (1) empirical study (1) instruction following (1) direct preference optimization (1) unsupervised learning (1) off-policy learning (1) semantic meaning (1) generative model (1) model generalization (1) model compression (1)