How does the brain extract relevant visual features from the rich, dynamic visual input that typifies active exploration, and how does the neural representation of these features support visual navigation?
Marius Schneider
(he/him)
Postdoctoral Scholar
Institute for Collaborative Biotechnologies
University of California, Santa Barbara
Marius Schneider is a Postdoctoral Scholar in the Bionic Vision Lab at UC Santa Barbara and recipient of the 2026–27 James V. & Beverly R. Zaleski Discovery Award in Robotics and Computing. He completed his doctoral research in systems neuroscience at the Ernst Strüngmann Institute in Frankfurt, in affiliation with the International Max Planck Research School for Neural Circuits, and defended his PhD with highest honors at Radboud University Nijmegen in May 2024.
Marius studies how biological and artificial visual systems process information during active behavior. His work combines dynamical systems, spiking neural networks, closed-loop NeuroAI, and large-scale neural data to understand how movement and brain state shape visual computation. At UCSB, he develops digital-twin models of active mouse vision, leads Mouse-vs-AI, and is launching a new research direction on neuromorphic safety computations for low-power edge systems.
Outside the lab, Marius enjoys running, working out at the gym, spending time at the beach, traveling, and listening to music.
- 3201C BioEngineering
- marius_schneider@ucsb.edu
Honors & Awards
- James V. & Beverly R. Zaleski Discovery Award in Robotics and Computing, UCSB (2026)
- EBBS Young Investigator Award (2024)
- Travel Grant for CNS 2019, Barcelona, Spain
- Travel Grant for Neural Dynamics Summer School 2018, Bristol, UK
Education
-
PhD in Neurophysics, 2024
Radboud University, Nijmegen, Netherlands
-
MS in Physics, 2019
Goethe University, Frankfurt am Main, Germany
-
BS in Physics, 2016
Goethe University, Frankfurt am Main, Germany
Project Lead
Publications
Gaze shifts in freely moving mice comprise distinct head-eye coordination motifs
We found that freely moving mice use multiple structured head-eye coordination motifs to shift gaze, including a Head-with-Eye motif that appears to reflect active visual orienting during natural behavior.
Yuchen Hou, Marius Schneider, Jhoseph Shin, Cristopher M. Niell, Michael Beyeler bioRxiv
(Note: YH and MS contributed equally to this work.)
Beyond neural activity prediction: Probing latent representations in mouse V1 digital twins
We introduce a multi-level evaluation framework for digital twins of mouse V1 that links neural-prediction accuracy to probe decodability, latent-unit tuning, and hidden-population geometry.
Adriano Lima, Yuchen Hou, Michael Beyeler, Marius Schneider arXiv:2605.23122
Visual robustness and neural alignment in a shared foraging task: The Mouse vs. AI benchmark
We introduce Mouse vs. AI, a public benchmark suite that unifies visual robustness, embodied foraging behavior, and neural alignment by evaluating artificial agents and mice in the same naturalistic 3D task.
Marius Schneider, Joe S. Canzano, Yuchen Hou, Jing Peng, Anjali Deepu, Utsab Karan, Phu-Hoa Pham, Tran Chi Nguyen, Dao Sy Duy Minh, Phu Quy Nguyen Lam, Trung-Kiet Huynh, Simone Azeglio, Spencer LaVere Smith, Michael Beyeler arXiv
