subspaces

My publications on Medium

Native Hierarchical and Compositional Representations with Subspace Embeddings. KDD 2026

Code for the paper Learning Visual-Semantic Subspace Representations

G-RANS is a research implementation of Generalizable Residual-Aware Neural Solvers for Sparse Systems (ICML 2026). It solves sparse linear systems from PDE discretizations with residual-aware neural subspace corrections.

A 15 part study guide for undergraduate students learning Linear Algebra