I have updated my preprint on how distorted grid-cell firing patterns might deform our memory on bioRxiv!
In this project, we investigate how distortions of grid-cell firing patterns could affect human spatial memory. Participants navigated a square and a trapezoidal environment using an immersive VR setup. They learned the positions of objects in both environments. Consistent with our predictions based on the distorted firing patterns of grid cells in rats, memory for positions was less precise in the trapezoid than in the square, an effect that was most pronounced in the narrow end of the trapezoid. In the revised version of the preprint, we use the eigenvector of the successor representation, which capture the distortions of grid patterns in a trapezoid, to demonstrate that spikes sampled from distorted grid patterns indeed carry less precise spatial information.
In additional memory tests, we asked participants to estimate the distances between objects - a task grid cells are thought to be important for. Participants estimated identical distances to be different between the square and the trapezoid and between the broad and narrow end of the trapezoid. Again, the findings were in line with the analyses of the distorted successor representation grid patterns, which revealed differences in the packing of grid fields.
If you want to find out more you can find the preprint on bioRxiv. Let me know if you have questions or comments!