Computation at criticality or edge of chaos

Establish whether computation in reservoir computing is most effective near a critical point or at the edge of chaos by rigorously comparing performance regimes and identifying the mechanisms that confer computational advantage.

Background

A recurring hypothesis in complex systems is that computation is optimal near critical points or at the edge of chaos. The paper highlights that, despite extensive discussion, it remains unresolved whether reservoirs achieve their best performance in such regimes. Determining this would shape guidance for tuning physical reservoirs and interpreting their dynamics.