Lev Petersen

I have always been interested in bounding problems. My life's passion thus far has been finding the edges, limits, and forms of things in domains of increasing abstraction. I have worked on and studied, in no particular order: reinforcement-learning systems, both continuous and discrete, with a particular interest in classical actor-critic research (world models optional!); metaprogramming, especially C++26's reflection paradigm and the future code-generation patterns it may enable; and stochastic processes, particularly actor models of markets.

I am interested in collaboration, though I am constrained in the scope of what I can contractually undertake. I am especially drawn to collaborations that make room for rigorous questions, clear interfaces, and ideas pursued all the way to their edges.

“There is a line somewhere in Wozzeck that translates out to, roughly, ‘The world is awful.’ Yes, I said to myself as I shot across the Bay Bridge not giving a fuck how fast I drove, that sums it up. That is high art: ‘The world is awful.’ That says it all. This is what we pay composers and painters and the great writers to do: tell us this; from figuring this out, they earn a living. What a masterful, incisive insight. What penetrating intelligence. A rat in a drain ditch could tell you the same thing, were it able to talk. If rats could talk, I’d do anything they said.”