Machine Learning
PINNs, Neural Operators and Can Transformers Be Represented as Lower-Dimensional Operators?
What a neural network actually does when it learns physics. From differential equations to Fourier Neural Operators, inverse parameter discovery, and the operator-theoretic foundations of transformer attention.
By Sankalp ChudmungeAugust 26, 2026
Machine Learning
From Numbers to Knowledge
What a machine actually does when it trains. From tensors to transformers, through norms, filters, and attention.
By Sankalp ChudmungeAugust 17, 2026
Linear Algebra & Numerical Methods
Why Power Iteration Works
Power iteration turns eigenvector discovery into repeated matrix multiplication. By leveraging spectral gaps, non-dominant components exponentially decay into numerical noise.
By Sankalp Chudmunge — Engineering LeadAugust 13, 2026