About me

Hi! I’m Duncan Soiffer, a 2024 - 2025 (December) Master’s in Machine Learning student at Carnegie Mellon University. Previously, I was an undergraduate at Worcester Polytechnic Institute, where I majored in computer science and math.

I am broadly interested in all aspects of machine learning, but especially interested in areas using theoretical insights to design (or simply inspire) methodology for important scientific or empirical applications, especially in representation learning and deep reinforcement learning.

Fun fact: my Erdős number is 2!

Current Projects

Currently, I am working in Professor Pradeep Ravikumar’s lab on weakly supervised representation learning with the ultimate goal of predicting harmful drug pair interactions with only single-drug intervention data.

On the side, I’m also experimenting with exploring inductive bias and the solution space of representation learning methods through a novel (but as-of-yet unproven) lens, a project examining the feasibility of learning how much to plan based on the state in model-based reinforcement learning, and a project designing search algorithms for a constructive mathematical proof.

Recent Projects

In Professor Virginia Smith’s lab, I led a project on extending agreement-based cascading to generative language tasks (where agreement is difficult to quantify).

Previous Projects

I have previously worked on reinforcement learning, applied ML in science, agentic AI, and a cubesatellite. Out of all of my projects so far, I’m most proud of my Major Qualifying Project (undergraduate thesis) at WPI. Along with fellow undergraduates Andrew Salls and Chase Miller, and supervised by Professors Daniel Reichman, Gábor Sárközy, and George Heineman, we improved on almost all lower bounds known for the Chvátal–Sankoff constants (click here for our paper) and additionally created a website to help teach and explore the Longest Common Subsequence problem more generally.

Other

I also like to play chess and ride my bike. I am also a very very amateur music composer, and have composed music for Superfly VR. If you have any suggestions on good DAWs and how to learn them, please let me know :)