Alexander Mathis

Alexander Mathis is an Assistant Professor at the École Polytechnique Fédérale de Lausanne (EPFL) in the Brain Mind Institute (School of Life Sciences). His research bridges computational neuroscience and machine learning to decode sensorimotor behaviors and develop AI tools for behavioral analysis.

• Pure Mathematics MSc, Ludwig Maximilians University Munich
• PhD in Computational Neuroscience, Ludwig Maximilians University Munich

His work focuses on understanding how the brain generates behavior through computational models and algorithms. Key contributions include the DeepLabCut toolbox, hBehaveMAE, and other frameworks for pose estimation, action segmentation, and brain-inspired AI. His group also explores proprioception, motor control, and neural coding theories.

Recent publications include unsupervised hierarchical behavior modeling via masked autoencoders (ECCV 2024), synthetic basketball benchmarks (Shot7M2), and extensions of BABEL into hBABEL. These works highlight his interest in temporal hierarchies in behavior and scalable AI solutions for neuroscience.

Scientific Honors

• Robert Bing Prize (2024)
• Eric Kandel Young Neuroscientists Prize (2023)
• Frontiers of Science Award (2023)
• Marie Skłodowska-Curie Postdoctoral Fellowship
• Studienstiftung des deutschen Volkes scholarship

He advises PhD candidates in neuroscience and life sciences, leads courses on brain-like computation and software engineering for life sciences, and collaborates across AI4Science initiatives. His group actively participates in competitions, such as NeurIPS' MyoChallenge, where brain-inspired reinforcement learning algorithms have won awards.