A new approach to interpretability: round-trip neural network compilation-decompilation

After learning about hyperdimensional computing I ended up making a programming language that is quite different from programming languages I know of.Sutra is a typed, GPU-native programming language I have been building. Its values are vectors and its programs compile to tensor-op graphs, the same kind of fused tensor computation a small neural network runs as. The paper is at arXiv:2605.20919 and the compiler is on GitHub.This post is about one specific property of that setup, which I will cal...