Luke T. Coddington

Papers

Total Citations

130

H-Index

About

Luke T. Coddington’s research lies at the intersection of computational neuroscience and reinforcement learning, with a primary focus on how mesolimbic dopamine shapes adaptive behavior. His major contribution centers on demonstrating that dopamine does more than signal reward prediction error—it dynamically adjusts the rate at which organisms learn from their own actions. In his highly cited 2023 paper (122 citations), Coddington showed that dopamine activity in the mesolimbic pathway modulates the balance between policy learning and value learning, effectively allowing the brain to optimize both behavioral strategies and reward predictions in real time. This work bridges a critical gap between artificial intelligence algorithms and neural mechanisms, offering a biologically grounded framework for understanding how agents—biological or artificial—can efficiently learn from sparse feedback. His 2021 preprint (8 citations) laid the foundational theory for this idea, establishing him as a rising voice in the field. Coddington’s research has profound implications for both AI development and the neuroscience of motivation, learning, and decision-making, making his work essential reading for students and researchers exploring how the brain computes adaptive action.