I am… a final-year PhD student at MIT with Antonio Torralba and Phillip Isola. I work on learned representations, decision-making, and reasoning.

Research. I am interested in the two intertwined paths towards generalist agents: (1) agents trained for multi-task (2) agents that can adaptively solve new tasks. My research learns structured representations that aggregate and select information about the world from various data sources, improving efficiency and generality of decision-making agents.

Outside research, I worked on building PyTorch (ver. 0.2 to 1.2), developing MIT’s Deep Learning course, coding open-source projects , mentoring SGI students (blog), holding pro bono office hours (book me!), brewing coffee ☕, travelling, and with 😸😼.

Email: tongzhou _AT_ mit _DOT_ edu

Selected Publications *indicates equal contribution; full list

Representations of generalist agents

What is machine learning really learning?

The Platonic Representation Hypothesis [ICML 2024 (Position Paper)][Project Page] [arXiv] [Code] Minyoung Huh*, Brian Cheung*, Tongzhou Wang*, Phillip Isola*	Hypothesis Strong AI models converge to a representation that fits reality, regardless of data, objective, or modality. rephrased; see paper for evidences and more.
Optimal Goal-Reaching Reinforcement Learning via Quasimetric Learning [ICML 2023][Project Page] [arXiv] [Code] Tongzhou Wang, Antonio Torralba, Phillip Isola, Amy Zhang	Quasimetric geometry + Push apart s_start & goal while keeping local dists = Optimal Goal-Reaching Agents
Denoised MDPs: Learning World Models Better Than The World Itself [ICML 2022] [Project Page] [arXiv] [code] Tongzhou Wang, Simon S. Du, Antonio Torralba, Phillip Isola, Amy Zhang, Yuandong Tian
On the Learning and Learnability of Quasimetrics [ICLR 2022] [Project Page] [arXiv] [OpenReview] [code] Tongzhou Wang, Phillip Isola
Learning to See by Looking at Noise [NeurIPS 2021] [Project Page] [arXiv] [code & datasets] Manel Baradad*, Jonas Wulff*, Tongzhou Wang, Phillip Isola, Antonio Torralba
Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere [ICML 2020] [Project Page] [arXiv] [code] Tongzhou Wang, Phillip Isola	# bsz : batch size (number of positive pairs) # d : latent dim # x : Tensor, shape=[bsz, d] # latents for one side of positive pairs # y : Tensor, shape=[bsz, d] # latents for the other side of positive pairs def align_loss(x, y, alpha=2): return (x - y).norm(p=2, dim=1).pow(alpha).mean() def uniform_loss(x, t=2): return torch.pdist(x, p=2).pow(2).mul(-t).exp().mean().log() PyTorch implementation of the alignment and uniformity losses
Dataset Distillation [Project Page] [arXiv] [code] [DD Papers] Tongzhou Wang, Jun-Yan Zhu, Antonio Torralba, Alexei A. Efros