Adam S. Lowet
I like using machine learning to understand brains and using neuroscience to understand machines.
Hi there! I'm a neuro-AI researcher based in the San Francisco Bay Area interested in a broad range of topics, including systems neuroscience, neuroethology, reinforcement learning, and mechanistic interpretability. I'm currently a postdoctoral researcher at UC Berkeley in the lab of Michael Yartsev, where I combine cutting-edge ML techniques with the extraordinary behavioral repertoire of bats to answer fundamental questions about decision-making in social and naturalistic contexts. Before that, I completed my Ph.D. in Neuroscience at Harvard, where I worked with Naoshige Uchida on the neural circuit basis of distributional RL. I also spent a brief stint conducting research with Kola Ayonrinde on mechanistic interpretability of large language models, using novel variants of sparse autoencoders to understand LLM features with differing levels of sparsity and abstraction. Feel free to reach out to me with any comments or requests for collaboration!
Recent Updates
- November 2025 Pleased to have received the Nemko Prize in Cellular or Molecular Neuroscience from the Society for Neuroscience.
- November 2025 Very much enjoyed giving an (opinionated) overview of the intersection of RL and neuroscience to the Sensorimotor AI Journal Club as part of their RL Debate Series!
- July 2025 Thrilled to be joining the NeuroBat Lab as a postdoctoral researcher!
- June 2025 Honored to receive the UZH Award for Research in Brain Diseases for the "pioneering discovery that the striatum can encode complete probability distributions" during my PhD.
- February 2025 Published "An opponent striatal circuit for distributional reinforcement learning" in Nature. (Press Release)
- December 2024 Received the Bowdoin Prize for my essay on risk and reward learning in brains and machines.
- October 2024 Succesfully defended my Ph.D. thesis at Harvard.