I am an undergrad Data Science student at UCSD with a Math minor. I am broadly interested in generalization and robustness, whether that be for audio data, graph learning, or algorithms, as well as designing inductive biases for neural network architectures that improve performance over traditional neural networks.

Recently, my research has focused on neural algorithmic reasoning and LLMs for algorithm discovery.

Currently I am working on a couple of cool things with cool people!

I am collaborating with:

• Prof. Yusu Wang on neural algorithmic reasoning and reinforcement learning for neural combinatorial optimization

• Prof. Andrea Chiba on building audio pipelines and models for stress detection in children

I have also worked on building benchmarks for graph information retrieval as a Ledell Scholar at UCSD, building knowledge graphs for pharmacology databases with the National Nanotechnology Initiative, and non-Euclidean models with Prof. Rex Ying at Yale.

Other related interests are geometric deep learning, equivariant learning, optimization theory, and any subfield of math!

Currently, I am diving more into mathematical analysis and group theory.

You can probably find me at the lab, at a cafe, or in my bed!

Feel free to reach out via email at prs007@ucsd.edu or pranavsingh.us@gmail.com.