Postdoctoral Scholar
Gallant Lab, University of California, Berkeley
t.zhang [at] berkeley.edu
I use dynamic neuroimaging experiments and use advanced computational methods to understand the cognitive and neural basis of complex, naturalistic human behavior.
My research focuses on spatial navigation, collaborative problem solving, and dynamic multiagent interactions.
Real-world experiences emerge out of a continuous perception-cognition-action loop that cannot be captured by classical experimental paradigms. To enable ecologically valid experiments, I built a naturalistic neuroimaging paradigm that immerse participants in dynamic virtual worlds, and utilitizes the capabilities of modern game engines to extract features for analysis.
The brain is selective for complex, high-dimensional combinations of features. These tuning properties are difficult to understand intuitively. We use a VAE to decode these tuning …
Visual-semantic representations in the human cortex differs between tasks, demonstrating that the brain reconfigures its tuning accoring to task demands. These results highlight …
The first quantitative map of the cortical regions that are engaged as people actively navigate. Active navigation engages multiple regions across the visual, parietal, and …