About Ferenc

My name is Ferenc Huszár. I am a machine learning researcher, I did my PhD in Cambridge with Carl Rasmussen, Máté Lengyel and Zoubin Ghahramani. I'm interested in probabilistic inference, generative models, unsupervised learning and applying deep learning to these problems.


I started this blog some time ago, in its current form, inFERENCe really started in 2015 with this post about probabilistic interpretation of autoencoders. I have returned to machine learning research after a 3-year gap spent as a data scientist. In those three years, the deep learning revolution happened, so I had to play catch up.

Initially, these blog posts helped me understand the body of literature I have missed: generative adversarial networks, variational autoencoders, etc. Nowadays, I continue reading and writing about current papers, trying to place them into my favourite framework of variational bounds, scoring rules, KL-divergences and information theory.

Twitter Cortex, etc.

I work at Twitter Cortex on unsupervised learning for visual data. We are building some awesome products that will make video a better experience for more people, using AI technology. My London-based team works on a number of research areas ranging from low-level vision to generative modeling and compression.

I ended up at Twitter when they acquired our startup, Magic Pony Technology, and as it turns out, it is a brilliant place to work. Prior to Magic Pony I worked as part of the investment team at Balderton Capital, an early-stage venture capital firm in London (hence the .vc domain name).

You can also find me on twitter @fhuszar.