I studied physics for a while, and now I study machine learning. Part of that old flame still burns in me, though, and I'm interested in how to put the math / physics into the machine. I'm also interested in Bayesian machine learning, continual learning, and model compression.
I consider myself an animal of the left, but I don't love everything coming out of it at the moment. I still think it's far better than whatever the other side are offering. This blog is not for politics — it's for free association; n.b., that's the grandiose term I use for when I'm talking without thinking.