Daniel Zhu

Hi!

I'm a ML Alignment and Theory Scholar (MATS), where I research multi-agent misalignment with Judy Shen, Erik Jones, and Jascha Sohl-Dickstein.

If you'd like to connect, feel free to reach out!

About

I worked as a Data Scientist at Pendulum, where I developed LLM-powered product features, fine-tuned BERT models for high-risk community classification, designed embeddings for social media channels, and architected an adaptive data ingestion system.

I graduated from the University of Washington in 2021 with a BS in Computer Science and a minor in Entrepreneurship. During my time at UW, I co-founded VerbalEyes, a startup focused on making digital video content accessible through automated video narration. I was also an inaugural entrepreneurship fellow at Madrona Venture Labs.

Earlier on, I was a Software Development Engineer Intern at Amazon, where I implemented a distributed index for Amazon's largest data lake; AI Research Fellow at Giving Tech Labs, where I constructed a social impact funding knowledge graph; and, a Teaching Assistant for CSE 446 Machine Learning at UW.