Efficient LLM Training & Inference Research on making large language models more efficient to train and deploy LLM Capabilities & Applications Research on understanding and extending the capabilities of large language models Deep Generative Models Research on improving and evaluating generative models, with focus on VAEs, GANs, and autoregressive models Semantic Parsing & Code Generation Research on improving semantic parsing and code generation through neural approaches Adversarial Learning and Robustness Research on adversarial robustness and adversarial learning approaches Gaussian Process Research A series of works on scaling and improving Gaussian Process models
