Charlie Yichuan Tang

2025-present

Stealth

2020-2025

I was a VP in D. E. Shaw & Co., in the systematic equities as well as the machine learning ventures groups, with a focus on alpha generation and incorporating AI/ML into various quant strategies.

2019-2020

Senior Research Scientist in the AI/ML Research division. Worked on temporal predictions and multi-agent robust Reinforcement Learning with applications to robotics and autonomy.

2016-2019

Research Scientist in the Special Projects Group (SPG). Worked on perception, prediction, planning, and RL for autonomous systems.

2015-2016

After graduating, I was the president and co-founder of a startup, Perceptual Machines, with my co-advisor Ruslan Salakhutdinov and lab-mate Nitish Srivastava.

We focused on deep learning for robotics (perception and prediction) including developing a deep learning training platform (this was all prior to Tensorflow and Pytorch), and inventing novel model architectures for autonomy.

After working closely with Apple (SPG) for a year, we were acquired by Apple in 2016.

2010-2015

PhD in AI, University of Toronto. Advised by Geoffrey Hinton and Ruslan Salakhutdinov.

My PhD thesis revolved around latent variable generative models in AI using energy-based training, which lead to interesting results in representation learning, robust perception, and invariances, despite limited on compute (in the era before diffusion and VLMs).

I researched visual attention (motivated by neuroscience) for invariant visual recognition before ViTs became the defacto standard in computer vision.

Motivated by learning multimodal distributions using only feed-forward MLPs, our stochastic neural nets research was prior to and related to the popular VAEs.

2008-2010

Masters in Computer Science from the University of Waterloo. I worked with Chris Eliasmith and was part of the Computational Neuroscience Research Group, including part of the research on a large-scale functional model of the brain.

2003-2008

Bachelors in Mechatronics Engineering and Robotics from the University of Waterloo.

Selected Publications

See full list of publications

Multiple Futures Prediction

Yichuan Charlie Tang, Ruslan Salakhutdinov. Neural Information Processing Systems. arxiv.org/abs/1911.00997 (NeurIPS 2019)

Worst Cases Policy Gradients

Yichuan Charlie Tang, Jian Zhang, Ruslan Salakhutdinov. Conference on Robot Learning. arxiv.org/abs/1911.03618 (CoRL 2019)

Towards Learning Multi-agent Negotiations via Self-Play

Yichuan Charlie Tang. Autonomous Driving Workshop, International Conference on Computer Vision. arxiv.org/abs/2001.10208 (ICCVW 2019) Talk

Relational Mimic for Visual Adversarial Imitation Learning

Lionel Blonde, Yichuan Charlie Tang, Jian Zhang, Russ Webb. Arxiv Preprint. arxiv.org/abs/1912.08444 (Arxiv 2019)

Learning Generative Models using Visual Attention

Yichuan Charlie Tang, Nitish Srivastava, Ruslan Salakhutdinov. Neural Information Processing Systems. arxiv.org/abs/1312.6110 (NIPS 2014, Oral)

Learning Stochastic Feedforward Neural Networks

Yichuan Charlie Tang, Ruslan Salakhutdinov. Neural Information Processing Systems. link (NIPS 2013)

Tensor Analyzers

Yichuan Charlie Tang, Ruslan Salakhutdinov, Geoffrey Hinton. International Conference on Machine Learning. link (ICML 2013)

Deep Learning using Linear Support Vector Machines

Yichuan Charlie Tang. Arxiv Preprint. https://arxiv.org/abs/1306.0239 (Arxiv 2013)

Deep Mixtures of Factor Analyzers

Yichuan Charlie Tang, Ruslan Salakhutdinov, Geoffrey Hinton. International Conference on Machine Learning. arxiv.org/abs/1206.4635 (ICML 2012, Oral)

Deep Lambertian Networks

Yichuan Charlie Tang, Ruslan Salakhutdinov, Geoffrey Hinton. International Conference on Machine Learning. arxiv.org/abs/1206.6445 (ICML 2012, Oral)

Robust Boltzmann Machines for Denoising and Recognition

Yichuan Charlie Tang, Ruslan Salakhutdinov, Geoffrey Hinton. IEEE Conference on Computer Vision and Pattern Recognition. link (CVPR 2012)

Multiresolution Deep Belief Networks

Yichuan Charlie Tang, Abdul-rahman Mohamed. 15th International Conference on Artificial Intelligence and Statistics. link (AISTATS 2012)

A Large Model of the Functioning Brain

Chris Eliasmith, Terry Stewart, Xuan Choo, Trevor Bekolay, Travis DeWolf, Yichuan Charlie Tang, Daniel Rasmussen. Science 30, November 2012. (Science)

Deep Networks for Robust Visual Recognition

Yichuan Charlie Tang, Chris Eliasmith. International Conference on Machine Learning. link (ICML 2010)

Bio

Charlie is an AI researcher, quant, entrepreneur and an angel investor. His research is focused on deep learning for prediction, planning, perception and reinforcement learning, with applications to agentic AI, robotics, and financial markets and quantitative trading. He is one of the few competitors to have reached the #1 ranking on Kaggle, a widely popular machine learning competition platform. Charlie obtained his PhD in 2015 in Machine Learning from the University of Toronto. His thesis focused on various aspects of Deep Learning technology. Charlie also holds a Bachelors in Mechatronics Engineering and Masters in Computer Science from the University of Waterloo. Charlie is also a Canadian national chess master, and a two time (2001, 2002) high school chess champion of the great state of Ohio.