About

I'm a Research Scientist at Google DeepMind, where I work on reinforcement learning — from foundational questions about objectives beyond a single scalar reward, to applying RL to mathematics, games, and reasoning in large language models.

I completed my PhD at the Technion — Israel Institute of Technology, advised by Prof. Shie Mannor, working on deep reinforcement learning, interpretability, and hierarchical RL. Since joining DeepMind, my research has spanned general value functions and convex MDPs, meta-learning, and diversity-seeking RL algorithms.

More recently I've been applying these ideas beyond classic RL benchmarks: AlphaProof, a system that reached IMO silver-medal level at theorem proving; generating novel chess puzzles and studying stylistic diversity in superhuman chess and Go engines; and combining RL with Gemini to design origami crease patterns.