Interactive 3D explainer for programmable nonlinear waveguide
Published:
We have published an interactive 3D explainer of the programmable nonlinear waveguide work: link.
Blog posts
2026
2022
How a single photon can induce 100-photon-worth of radiation pressure
Published:
I have recently got a chance to read this seminal paper by Yakir Aharonov et al., where a major surprise of the work is that, for a certain construction of the spin measurement of spin-1/2 particles, the apparent readout of the spin of a single particle can be arbitrarily larger than 1/2 (e.g., 100). While their main results are provided in a generic language, the thought experiment considered in the paper is a somewhat involved Stern-Gerlach experiment using electrons, which I struggled to digest as a one from nonlinear optics. The aim of this post is to show a simple nonlinear-optical reformulation of their thought experiment, hopefully providing useful intuitions to those with quantum and/or nonlinear optics background. Specifically, we construct a concrete example where single-photon can induce 100-photon-worth of radiation pressure.
SDE simulation with complex non-diagonal noise on DifferentialEquations.jl
Published:
The aim of this post is to keep a note on a pitfall that I felt in when implementing a certain type of stochastic differential equation (SDE) using DifferentialEquations.jl, recapping the discussions that took place here.
2021
On Hermitian conjugation of time-ordered unitary
Published:
Taking Hermitian conjugation of an operator is an everyday task in quantum mechanics, while more care is needed when handling operators with some nontrivial structures. Here, this post addresses a pitfall that I recently fell in in an attempt to take a Hermitian conjugate of a unitary involving time-ordering. While the content of this post might be totally obvious for some, I have still managed to trick non-trivial number of people with years of experience in quantum mechanics with this problem.
Handling LAPACK exceptions in Julia SVD
Published:
I have run numerical simulations that involve a large number of singular value decompositions (SVDs) using Julia, where I noticed that SVD algorithm fails on a few specific matrices out of several millions. While the issue can be economically circumvented by switching SVD algorithms whenever they fail, exact cause of the error has not been clear yet.