hi, I am Anshuman

I like tech and am an ambivert geek. Currently, I'm working on evals and benchmarks to help make AI safer, with a focus on cybersecurity evals, AI alignment, AI welfare, and RFM (robotic foundation model) safety.

I've been building robots since middle school in boarding school and am now also poking at robotics benchmarks and evals — currently playing with models and experimenting with SO101 and other robotic arms. More here ↗


Experience