My Work for Wolfram Research (Almost)

I was accepted into the Wolfram Research Physics Program to have my research advised under Stephen Wolfram, but unfortunately had to decline due to work conflicts. We might be collaborating in the future but for now I’ll continue independently.

My research there would’ve been optimal annealing control of electronic oscillators for combinatorial optimization.

Annealing heuristics bother me. Can we derive better protocols from first principles of statistical (nonequilibrium) physics? There has been recent work on a formalism for optimized annealing but it’s application towards realistic multivariable optimization is unclear.

I built an analog circuit for an experimental testbed of this theory. How should an external synchronization be controlled over time to best guide the circuit dynamics to towards global minima? Can we design feedback to avoid this altogether? Decomposing the analog trajectories may help explore transition behavior during minimization of the Lyapunov and objective functions as the synchronization is varied. This works aims at developing efficient and scalable optimization hardware and software that utilizes inherit dynamics of stochastic (electronic) systems.