Riley Nerem
Neural algorithmic reasoning
Robert Riley Nerem
PhD student in Data Science · UC San Diego
I work on neural algorithmic reasoning, studying how neural networks learn and use algorithmic structure. My research focuses on generalization on graphs, discrete and combinatorial optimization, and designing neural networks for specific applications with an algorithmic inductive bias.
I am graduating soon and currently looking for research-focused opportunities starting Fall 2027!
I go by Riley.
Research interest: Neural Algorithmic Reasoning
01
Generalization on graphs
Understanding when neural networks generalize to larger graphs and under distribution shifts.
02
Combinatorial optimization
Developing learning methods for optimization over combinatorial structures such as graphs, permutations, and trees.
03
Algorithmic alignment
Designing neural architectures for specific applications with an inductive bias toward algorithmic behavior.