Keynotes | K&Ts | GACs | Talks | Posters | Search

Poster B in Poster Session B: Tuesday, August 4, 2:00 – 3:45 pm, Kimmel Center, Shorin & Rosenthal Rooms

Spatial uncertainty underlies geometric distortions in navigation and grid cell tuning

Yul HR Kang¹, Daniel Wolpert², Máté Lengyel³; ¹Korea Advanced Institute of Science & Technology, ²Columbia University, ³University of Cambridge

Presenter: Yul HR Kang

Navigation requires estimating one’s location from noisy sensory inputs. While environmental geometry is known to distort both navigational behavior and grid cell responses, these effects have been explained piecemeal. Here we propose a unifying principle: the brain maintains and updates uncertainty about spatial location, and this uncertainty shapes both behavior and neural coding. We introduce the Bayesian Image-computable Observer for Navigation (BION) that integrates visual and self-motion inputs to infer a posterior over location. Crucially, spatial uncertainty is structured by environmental geometry due to retinal projection, yielding systematic inhomogeneities in information. These in turn explain human homing across trapezoidal, stretched, and compressed environments. We further show that encoding this uncertainty in neural population codes predicts rat grid field size, anisotropy, rescaling, and boundary-tethering under analogous manipula- tions. Unlike prior models that treat distortions as epiphenomena, our framework interprets them as optimal adaptations to uncertainty. These results suggest that spatial uncertainty is not incidental but fundamental to navigation, providing a normative account linking perception, behavior, and neural representation.

Topic Area: Methods, Tools, Theory & Neural Coding