Projects

How deep can LLMs learn to reason? Explore how logical expressiveness shapes reinforcement learning, through interactive proofs and experimental results.