Attractor-Based Grid Cell Models and Aliasing Trade-offs

Investigate whether replacing the analytically constructed grid-cell signals with an attractor-based grid-cell model, particularly one with greater biological plausibility, introduces different trade-offs between spatial-aliasing mitigation and representation stability.

Background

The paper demonstrates that analytically constructed grid-cell signals substantially reduce spatial aliasing in BVC-driven place representations across open, symmetric, and maze-like environments. However, the implemented grid cells are not generated by a biologically grounded dynamical mechanism; they are constructed directly using cosine interference and obstacle-aware masking. The authors identify the use of an attractor-based grid-cell model as a concrete direction for future investigation, with the unresolved issue being whether increased biological plausibility would alter the observed balance between aliasing reduction and the stability of spatial representations.