Robert Joseph
I am currently a Research Intern at Amazon in NYC with the RL Team and a 3rd-year Ph.D. student in Math & CS at Caltech, advised by Professor Anima Anandkumar, leading our lab's AI4Math effort. My research is on AI for Math, Science, Code, and formal verification. I am gratefully supported by the Caltech Graduate Fellowship, Harmonic AI, and DARPA expMath.
Previously, I graduated from the University of Alberta in Honors Math & Computer Science with the Dean's Silver Medal in Science. I worked on reinforcement learning at the RLAI Lab, led by Professor Richard Sutton, with Professor Martha White and Adam White, and did internships and research with Google AI, Amii, Microsoft Research, PIMS, and AMI.
I am also interested in giving back to the community by teaching and mentoring students and sharing my love for AI4Science & Math. I previously co-led the ML Theory Reading Group at Cohere for AI. Feel free to contact me if any of these interests align with your research or if you have any questions. I am always open to new collaborations.
Latest News
- September 2026 - Gave a talk on Verified Scientific Machine Learning in Lean at IPAM, UCLA.
- September 2026 - Released FloatLib, our verified floating-point arithmetic library in Lean.
- September 2026 - Released a new stability framework for singularity formation in the 3D Euler equations.
- September 2026 - Gave a talk at Amazon on TorchLean and formally verified neural networks.
- July 2026 - Released ITPEval for formal translation across four theorem provers.
- June 2026 - Presented QuantumLean-Bench at ICML AI4Physics.