PhD student, Department of Computer Science, University of Oxford
1 paper at NeurIPS 2025
We consider graph neural networks with mean aggregation and non-negative weights, showing which monotonic rules they can learn and constructing sound explanatory rules.
