Hello,

This is my first-ever blog post (as far as I remember really). I planned to start blogging in my freshman year, and it’s been almost three years since then. Planning fallacy, as Kahneman would say. Anyway, I should thank Google Summer of Code to finally make me do it (although technically so. Fret not, I will try to make these blogs as accessible as possible) as part of my upcoming summer project with NUMFOCUS’s QuTiP. Primarily, I will be working to make QuTiP auto-differentiable. A bit more about my project.

Modern Machine Learning libraries provide a way to exponentiate hamiltonian (matrices), but there is no smooth workflow that allows a user to take derivates of these exponentials (and physicists like that, you know). The project shall bring about autograd-type capabilties, but for exponentials of Hamliltonians. One obvious use-case of these auto-diff capabilities would be to the problems of hamiltonian Learning. Thus, the project shall also package gradient based optimizers to let the user perform tasks like unitary learning. Read the full details of the project in my proposal here.

I am very excited about the project, and so are my mentors. I will be posting bi-weekly blogs about the status of my project and my learning journey along the way. So, keep an eye if that’s something you would want to read about!