hi, i'm benjamin

I’m an engineer and researcher focused on building AI that’s aligned with human values. I am also working on research that uses machine learning to study policy outcomes in policing, transportation and government services. In a similar vein, I’m developing technical methods to empower local communities to advocate for cycling infrastructure.

I recently attended a programming residency at the Recurse Center, where I worked on projects related to technical AI safety and AI governance. While I was there I built a lot of things, including:

• An LLM tokenizer in Python that uses C extensions to optimize performance
• An implementation of GPT-2 where I only relied on research papers for reference
• A deepdream algorithm to visualize parameters of a CNN image classifier and developed methods to “blind” the model to specific objects using principles of mechanistic interpretability

You can e-mail me if you’d like to chat about new projects, cycling, music or ways to make the world a better place for all humans.

I like to learn things and tell other people about them

Read my latest article in Vital City where I highlight a chronic issue in New York City and how it’s emblematic of the challenges of crafting effective policies when underlying data is tainted. This piece is largely based on this research paper where I used machine learning to study how police respond to illegal parking complaints. It paints dire picture where the majority of cases are ignored and offers concrete solutions to make streets safer.

This built on earlier work I did using computer vision and official data to understand the chaotic nature of New York City streets.

I like to tinker

I was tired of checking my phone for subway departures, so I built an LED arrival board for my living room.