Guangyao (Stannis) Zhou

Research Scientist, DeepMind

stannis [AT] deepmind.com

About

I am a Research Scientist at DeepMind. My current research interests are in building compositional generative models, and the associated inference and learning problems.

Before DeepMind, I worked for 3 years as a Research Scientist at Vicarious AI (acquired by Alphabet). I obtained my Ph.D. in applied math at Brown University, advised by Prof. Stuart Geman, and my bachelor degrees from Peking University.

Publications

Most recent publications on Google Scholar.
indicates equal contribution.

3D Neural Embedding Likelihood for Robust Probabilistic Inverse Graphics

Guangyao Zhou, Nishad Gothoskar, Lirui Wang, Joshua B Tenenbaum, Dan Gutfreund, Miguel Lázaro-Gredilla, Dileep George, Vikash K Mansinghka

arXiv:2302.03744, 2023

Graph schemas as abstractions for transfer learning, inference, and planning

J Swaroop Guntupalli, Rajkumar Vasudeva Raju, Shrinu Kushagra, Carter Wendelken, Danny Sawyer, Ishan Deshpande, Guangyao Zhou, Miguel Lázaro-Gredilla, Dileep George

arXiv:2302.07350, 2023

Learning noisy-OR Bayesian Networks with Max-Product Belief Propagation

Antoine Dedieu, Guangyao Zhou, Dileep George, Miguel Lázaro-Gredilla

International Conference on Machine Learning (ICML), 2023

Space is a latent sequence: Structured sequence learning as a unified theory of representation in the hippocampus

Rajkumar Vasudeva Raju, J Swaroop Guntupalli, Guangyao Zhou, Miguel Lázaro-Gredilla, Dileep George

arXiv:2212.01508, 2022

PGMax: Factor Graphs for Discrete Probabilistic Graphical Models and Loopy Belief Propagation in JAX

Guangyao Zhou, Antoine Dedieu, Nishanth Kumar, Wolfgang Lehrach, Miguel Lázaro-Gredilla, Shrinu Kushagra, Dileep George

arXiv:2202.04110, 2022

Metropolis Augmented Hamiltonian Monte Carlo

Guangyao Zhou

Symposium on Advances in Approximate Bayesian Inference (AABI) 2022

Graphical Models with Attention for Context-Specific Independence and an Application to Perceptual Grouping

Guangyao Zhou, Wolfgang Lehrach, Antoine Dedieu, Miguel Lázaro-Gredilla, Dileep George

arXiv:2112.03371, 2021

Query Training: Learning a Worse Model to Infer Better Marginals in Undirected Graphical Models with Hidden Variables

Miguel Lázaro-Gredilla, Wolfgang Lehrach, Nishad Gothoskar, Guangyao Zhou,
Antoine Dedieu, Dileep George

AAAI Conference on Artificial Ingelligence (AAAI) 2021

Mixed Hamiltonian Monte Carlo for mixed discrete and continuous variables

Guangyao Zhou

Advances in Neural Information Processing Systems (NeurIPS) 2020
Extended abstract accepted as talk at PROBPROG 2020

A detailed mathematical theory of thalamic and cortical microcircuits based on inference
in a generative vision model

Dileep George, Miguel Lázaro-Gredilla, Wolfgang Lehrach, Antoine Dedieu, Guangyao Zhou

bioRxiv 2020.09.09.290601, 2020

Capacities and efficient computation of first passage probabilities

Jackson Loper, Guangyao Zhou, Stuart Geman

Phys. Rev. E 102, 023304, 2020

Base-pair ambiguity and the kinetics of RNA folding

Guangyao Zhou, Jackson Loper, Stuart Geman

BMC Bioinformatics, 20(1):666, 2019

Sparse feature selection by information theory

Guangyao Zhou, Stuart Geman, Joachim M Buhmann

IEEE International Symposium on Information Theory (ISIT), 2014

L1-graph construction using structured sparsity

Guangyao Zhou, Zhiwu Lu, Yuxin Peng

Neurocomputing, 120:441-452, 2013

Vitæ

Full Resume in PDF.

Acknowledgement

This website uses the website design and template by Martin Saveski