Marina Zhang

CV / LinkedIn / Google Scholar / GitHub

prof_pic.jpg

Hello! I am a software engineer on Google’s Security & Anti-Abuse Research Team, led by Elie Bursztein. My research interests lie at the intersection of AI and security. I am excited to start my PhD at Stanford in September 2024!

At Google, my research is primarily focused on building natural language processing (NLP) models and embeddings that are secure, efficient, and robust against adversarial attacks. I’m also interested in designing novel, machine learning-based solutions to tackle important security challenges.

Prior to joining Google, I graduated from MIT in 2021 with a double major in Computer Science and Engineering & Mathematics.

News

Jan 16, 2024 Our paper on RETSim, a lightweight text embedding for robust text similarity, has been accepted at ICLR 2024! Check out the paper on arXiv.
Dec 4, 2023 I published a Google Security blog post on the launch of our new RETVec-powered Gmail spam filter! The launch has been covered by Forbes, Hacker News, PCMag, and more!
Nov 29, 2023 We are excited to announce the open-source beta release of the UniSim: Universal Similarity package for fast fuzzy matching, search, and clustering using embeddings!
Nov 27, 2023 We officially released the open-source RETVec package on GitHub! The package contains an easy-to-use TensorFlow/Keras API to help you build better text models with RETVec.
Sep 22, 2023 Our paper RETVec: Resilient and Efficient Text Vectorizer was accepted at NeurIPS 2023!

Research

RETSim: Resilient and Efficient Text Similarity
Marina Zhang, Owen Vallis, Aysegul Bumin, Tanay Vakharia, and Elie Bursztein.
International Conference on Learning Representations (ICLR), 2024.
[paper] [code]

Generic Attacks against Cryptographic Hardware through Long-Range Deep Learning
Elie Bursztein, Luca Invernizzi, Karel Král, Daniel Moghimi, Jean-Michel Picod, and Marina Zhang.
Conference on Cryptographic Hardware and Embedded Systems (CHES), 2024.
[paper] [code]

RETVec: Resilient and Efficient Text Vectorizer
Elie Bursztein, Marina Zhang, Owen Vallis, Xinyu Jia, and Alexey Kurakin.
Advances in Neural Information Processing Systems (NeurIPS), 2023.
[paper] [code]

Machine Learning-based Clustering and Classification of Mouse Behaviors via Respiratory Patterns
Emma Janke, Marina Zhang, Sang Eun Ryu, Janardhan Bhattarai, Mary Schreck, Andrew Moberly, Wenqin Luo, Long Ding, Daniel Wesson, and Minghong Ma.
iScience (Cell Press) Vol. 25, 2022.
[paper]

Scratch-AID: A Deep-learning Based System for Automatic Detection of Mouse Scratching Behavior with High Accuracy
Huasheng Yu, Jingwei Xiong, Adam Yongxin Ye, Suna Li Cranfill, Tariq Cannonier, Mayank Gautam, Marina Zhang, Rayan Bilal, Jong-Eun Park, Yuji Xue, Vidhur Polam, Zora Vujovic, Daniel Dai, William Ong, Jasper Ip, Amanda Hsieh, Nour Mimouni, Alejandra Lozada, Medhini Sosale, Alex Ahn, Minghong Ma, Long Ding, Javier Arsuaga, and Wenqin Luo.
eLife Vol. 11, 2022.
[paper]