Li Kevin Wenliang

Understanding Prompt Tuning and In-Context Learning via Meta-Learning

Amortized Planning with Large-Scale Transformers: A Case Study on Chess

Language Modeling Is Compression

Near-Minimax-Optimal Distributional Reinforcement Learning with a Generative Model

Distributional Bellman Operators over Mean Embeddings

Neural Networks and Solomonoff Induction

Generative Reinforcement Learning with Transformers