| CARVIEW |
Select Language
HTTP/2 200
server: GitHub.com
content-type: text/html; charset=utf-8
last-modified: Tue, 27 May 2025 16:36:28 GMT
access-control-allow-origin: *
etag: W/"6835ea0c-194e"
expires: Mon, 29 Dec 2025 13:10:12 GMT
cache-control: max-age=600
content-encoding: gzip
x-proxy-cache: MISS
x-github-request-id: D571:328FD3:8DADA0:9EFAA8:69527B5A
accept-ranges: bytes
age: 0
date: Mon, 29 Dec 2025 13:00:12 GMT
via: 1.1 varnish
x-served-by: cache-bom-vanm7210063-BOM
x-cache: MISS
x-cache-hits: 0
x-timer: S1767013212.974269,VS0,VE213
vary: Accept-Encoding
x-fastly-request-id: f5b45c35b8e540899e5b79529032e8667939db32
content-length: 2788
Hattie Zhou's Homepage
Hattie Zhou
Graduate Student at Mila
Graduate Student at Mila
Email: hattie.zhou [at] mila.quebec
About
Good day! I am currently a research scientist at Anthropic, working on improving frontier capabilities of the model across both pretraining and post-training.
I am also a PhD student at Université de Montréal and Mila, where I am supervised by Hugo Larochelle and Aaron Courville.
During my PhD, I had the pleasure to intern at Apple MLR with Preetum Nakkiran, Josh Susskind and Samy Bengio. I was also a student researcher at Google Brain in 2022, working with Hanie Sedghi and Behnam Neyshabur on making large language models reason better.
Prior to Mila, I was a data scientist at Uber where I built econometric and machine learning models to measure and optimize marketing spend globally. I also had the fortune of doing research with wonderful folks at Uber AI Labs (now ML Collective). These folks took me in when all I knew about deep learning was having done Andrew Ng's Coursera course and all I knew about python was what I'd gleaned from the tensorflow tutorials, proving that anyone can do research with the right amount of gentle guidance (and compute...). Before getting into the world of machine learning, I was an undergraduate at the Ivey Business School in Canada, where I studied finance and entrepreneurship (if one could even study such a thing). Prior to Uber, I worked as a private equity analyst at Radar Capital, and then as an economic consultant at Cornerstone Research.
During my PhD, I had the pleasure to intern at Apple MLR with Preetum Nakkiran, Josh Susskind and Samy Bengio. I was also a student researcher at Google Brain in 2022, working with Hanie Sedghi and Behnam Neyshabur on making large language models reason better.
Prior to Mila, I was a data scientist at Uber where I built econometric and machine learning models to measure and optimize marketing spend globally. I also had the fortune of doing research with wonderful folks at Uber AI Labs (now ML Collective). These folks took me in when all I knew about deep learning was having done Andrew Ng's Coursera course and all I knew about python was what I'd gleaned from the tensorflow tutorials, proving that anyone can do research with the right amount of gentle guidance (and compute...). Before getting into the world of machine learning, I was an undergraduate at the Ivey Business School in Canada, where I studied finance and entrepreneurship (if one could even study such a thing). Prior to Uber, I worked as a private equity analyst at Radar Capital, and then as an economic consultant at Cornerstone Research.
Select Research
- Step-by-step diffusion: An elementary tutorial
Preetum Nakkiran, Arwen Bradley, Hattie Zhou, Madhu Advani.
Foundations and Trends® in Computer Graphics and Vision: Vol. 17: No. 1, pp 1-75
Paper
- What Algorithms Can Transformers Learn? A Study in Length Generalization
Hattie Zhou, Arwen Bradley, Etai Littwin, Noam Razin, Omid Saremi, Josh Susskind, Samy Bengio, Preetum Nakkiran.
ICLR 2024
Paper
- Teaching Algorithmic Reasoning via In-context Learning
Hattie Zhou, Azade Nova, Hugo Larochelle, Aaron Courville, Behnam Neyshabur, Hanie Sedghi.
NeurIPS 2022 MathAI Workshop
Paper | Blog
- Fortuitous Forgetting in Connectionist Networks
Hattie Zhou, Ankit Vani, Hugo Larochelle, Aaron Courville.
ICLR 2022
Paper | Code
- LCA: Loss Change Allocation for Neural Network Training
Janice Lan, Rosanne Liu, Hattie Zhou, Jason Yosinski.
NeurIPS 2019
Paper | Code | Blog
- Deconstructing Lottery Tickets: Zeros, Signs, and the Supermask
Hattie Zhou, Janice Lan, Rosanne Liu, Jason Yosinski.
NeurIPS 2019
Paper | Code | Blog
Talks
- I was on the Gradient Podcast!
- I had a fun conversation on the Generally Intelligent Podcast about my research.
- I made a brief appearance on Machine Learning Street Talk discussing our work on teaching algorithmic reasoning.
- Long time ago, I gave a talk at the Brains@Bay meetup about deconstructing lottery tickets - my first long-form research talk!
Here are some recordings of me discussing my research if you like these alternative formats:
Misc
- Feel free to ping me on Twitter, or drop me an Email!
Contact
Lastly, as a lover of dad jokes, I share a curated one below, which I will change periodically. Enjoy!
See here for a list of past jokes.
I just found out my boyfriend is a ghost. I should have known the moment he walked through the door.