Tyler Zhu Personal Website computer science, mathematics, and machine learning.

Welcome!

Hey there! I'm Tyler, a 5th-year Master's in EECS at the University of California, Berkeley advised by Jitendra Malik. Broadly, my main interests are mathematics and computer science, specifically in computer vision.

I am interested in creating intelligent and trustworthy systems that can interact and reason about our world. Specifically, my goal is to build vision systems that are efficient, robust, and intelligent. My long term goal is to create models with human alignment in mind, so that their behavior is both useful and safe. Currently, I work in the computer vision group with Karttikeya Mangalam on creating models which maintain meaningful, persistent memories of data over much longer contexts than currently offered by typical transformer-based approaches (around 4k tokens).

Previously, I worked on improving monocular depth estimation techniques with Alvin Wan in Joey Gonzalez's group in RISE Lab. Before that, I worked with Dan Hendrycks in Dawn Song's group on creating diverse augmentations with neural networks as well as establishing a widely used set of benchmarks for evaluating robustness of neural networks.

While at Berkeley, I was heavily involved in both teaching and outreach, which were some of the most rewarding activities I pursued. I was a TA for CS 70 three times, once as a head uGSI, and a reader for CS 174 once, creating a mass of resources for students. I was also (and still am) very involved in Machine Learning @ Berkeley, a student organization dedicated to creating a vibrant machine learning community at Berkeley. I served on the leadership team multiple times, and was president during the Spring 2021 semester. I helped organize many research projects and connections, and mentored many students in the organization.

Aside from academics, I play guitar and basketball and love to watch sports (go Giants and Warriors), i and I also enjoy music, food, and coffee to an unusual extent. If you're equally interested in any of the above, let's chat! My email is tyler [dot] zhu [at] berkeley [dot] edu.

You can find some various things that I work on around my website. Below are my coursework and some course notes I've taken, as well as some teaching resources I've made over the years. There's also my personal blog, which I'll start writing in... one day. Here are some pointers to things I think you'll find interesting:

  • My discussion sheets for CS 70 from Spring 2020, which are very popular among students. You can also find my evaluations here.
  • My course notes for EECS 126 and EECS 127.
  • My website for project 5 in CS 194-26, where I implemented a form of GCNet (from depth estimation) to do facial keypoint detection.

Coursework

Here's an abridged summary of the classes I have taken. Some of them have links to additional resources that I've created. The notes are written in LaTeX and live during lecture, so they may contain many errors. I try my best to correct them after the fact. My setup is inspired by that of Evan Chen's. You can find some of them in my public course notes repository.

Fall 2022

  • CS 285: Deep Reinforcement Learning

Spring 2022

  • CS 182: Designing, Visualizing and Understanding Deep Neural Networks
  • CS 184: Computer Graphics
  • CS 280: Computer Vision
  • History of Art C11: Introduction to Western Art: Renaissance to the Present

Fall 2021

Spring 2021

Fall 2020

  • CS 294-165: Sketching Algorithms
  • Math 104: Real Analysis
  • Math 250A: Groups, Rings, and Fields

Spring 2020

Fall 2019

  • CS 61C: Great Ideas of Computer Architecture
  • CS 189: Introduction to Machine Learning

Spring 2019

Fall 2018

Bio

Mar 2022 - Present CV Research w/ Karttikeya Mangalam, Prof. Jitendra Malik @ Berkeley AI Research
Sep 2021 - Mar 2022 CV Research w/ Alvin Wan, Prof. Joey Gonzalez @ RISE Lab
Jun 2020 - Aug 2020 Trading Intern @ Citadel Securities
Jan 2020 - June 2020 Trustworthy ML Research w/ with Dan Hendryks, Prof. Dawn Song @ Berkeley AI Research
May 2019 - Aug 2019 Software Engineering Intern @ Google