Aidan Swope is a research scientist in California with 11 years of experience building neural networks that aim to reason more like humans. He focuses on scalable automated reasoning via language models, reinforcement learning, human preference modeling, and discrete search methods applied to theorem proving and circuit optimization. His selected work includes LeanDojo for retrieval-augmented theorem proving and CircuitVAE for latent circuit optimization, and he has probed why networks struggle with math through collaborations at Caltech and NVIDIA. At Harmonic he develops LM + search + RL systems to solve formal math problems at competition and research level, blending deep theory with practical engineering. Known for “making big, wiggly functions and taming them with optimization,” he brings both hands-on ML systems experience and a taste for elegant, principled solutions.
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.