Publications

Shane's research spans reinforcement learning, generative modeling, probabilistic inference, and large language model reasoning — often bridging these areas to develop algorithms that are both theoretically grounded and practically scalable. Early work focused on sample-efficient deep RL and Bayesian methods; later work extended into offline RL, robotic control, and LLM reasoning, culminating in contributions to frontier models like Gemini.

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015