About Me

I am a PhD student at Berkeley AI Research (BAIR), advised by Jitendra Malik. I am also affiliated with the NVIDIA GEAR group led by Jim Fan and Yuke Zhu. During the early years of my PhD, I had the fortune of being mentored by Alexei Efros, who not only guided my academic growth but also cultivated my culinary, literary, and cultural taste.

Previously, I obtained my BSc and MEng from MIT EECS, under the supervision of Phillip Isola and Antonio Torralba. Before transferring to MIT, I was an undergraduate student at The University of Tokyo. I spent the past summers interning at DeepMind (2021, 2019), Facebook (2018), and Google (2017, 2016).

I am currently building real-world robot systems that can achieve dexterous low-level skills through end-to-end learning.

Research Interests

I work on applying learning methods to enable low-level dexterous manipulation skills and whole-body control on real robots, leveraging data from both the real world and simulators. Feel free to reach out if interested in research discussion / collaboration!

Learning Methods reinforcement learning, imitation learning, sim-to-real, generative models

Hardware Systems bimanual arms with dexterous hands, humanoids

Publications

HOVER: Versatile Neural Whole-Body Controller for Humanoid Robots
Tairan He*, Wenli Xiao*, Toru Lin, Zhengyi Luo, Zhenjia Xu, Zhenyu Jiang, Jan Kautz, Changliu Liu, Guanya Shi, Xiaolong Wang, Linxi "Jim" Fan†, and Yuke Zhu†.
arXiv 2024.

Learning Visuotactile Skills with Two Multifingered Hands
Toru Lin, Yu Zhang*, Qiyang Li*, Haozhi Qi*, Brent Yi, Sergey Levine, and Jitendra Malik.
arXiv 2024.

Twisting Lids Off with Two Hands
Toru Lin*, Zhao-Heng Yin*, Haozhi Qi, Pieter Abbeel, and Jitendra Malik.
CoRL 2024.

MIMEx: Intrinsic Rewards from Masked Input Modeling
Toru Lin and Allan Jabri.
NeurIPS 2023.

Learning to Ground Multi-Agent Communication with Autoencoders
Toru Lin, Minyoung Huh, and Phillip Isola.
NeurIPS 2021.

Geometry-Aware Modeling of Rigid Body Physics
Toru Lin*, Kexin Yi*, and Phillip Isola.
ICML 2020, Workshop on Object-Oriented Learning.

Visual Grounding of Learned Physical Models
Yunzhu Li, Toru Lin*, Kexin Yi*, Daniel Bear, Daniel LK Yamins, Jiajun Wu, Joshua B Tenenbaum, and Antonio Torralba.
ICML 2020.

Model-Based Planning with Energy-Based Models
Yilun Du, Toru Lin, and Igor Mordatch.
CoRL 2019.

Teaching

I served as a teaching / lab assistant for the following courses.

UC Berkeley CS 294-277 Robots That Learn (Fall 2024)
MIT 6.819/6.869 Advances in Computer Vision (Spring 2021)
MIT 6.036 Introduction to Machine Learning (Fall 2019)
MIT 6.863/9.611 Natural Language Processing (Spring 2018)

Miscellaneous

I speak English, Japanese, Mandarin, Cantonese and Singlish natively.
Outside of research, I enjoy reading philosophy, learning physics, and making art.