Kenneth Jabon

Boston, MA:

Open to relocation

I’m a CS/Photonics guy out of UIUC, and I’m deeply interested in applying deep reinforcement learning (RL) to challenging problems.
If you would like to contact me RE: employment, see links at the bottom.

See my projects!

I’ve applied RL to…

Photonic device design
Personal habit formation
Air hockey played by robots

… and I’ve implemented MuZero with efficiency improvements in Acme.

I’ve also done the following work, resulting in publications:

Automated high-volume manufacturing, and wafer-scale test, of photonic devices with high-precision robotics (at IPG and AP respectively)
Designed and taped out microring resonator “molecules” for photon pair generation. If you’re making a linear-optic quantum computer, I have a device for you. I was lucky to be advised by Miloš Popović.

Can RL be applied to automating these things? Several years ago, smart money would have said not yet. Now, robots are learning to walk in the real world in less than 20 minutes, agents are defeating grandmasters in Go, and training things as impressive as ChatGPT. I’m optimistic for this technology’s future.