Deqing Fu

transformer architecture (1) mathematical reasoning (1) self-supervised learning (1) in-context learning (1) question answering (1) convergence analysis (1) image synthesis (1) visual grounding (1) text-to-image generation (1) cross-modal transfer (1) semantic alignment (1) model interpretability (1) linear regression (1) newton's method (1) second-order optimization (1) vision language model (1) domain translation (1) multimodal large language model (1) mechanistic interpretability (1) sparse autoencoder (1)