xiaobo liang

non-autoregressive generation (3) test-time scaling (2) inference efficiency (2) machine translation (2) model compression (2) neural network (2) preference optimization (2) large language model (2) generative reward model (2) domain adaptation (2) reinforcement learning (2) masked language model (1) preference learning (1) language modeling (1) named entity recognition (1) kl divergence (1) mathematical reasoning (1) bidirectional attention (1) preference alignment (1) cross-modal transfer (1)