Search papers, labs, and topics across Lattice.
This study investigates the relationship between the spectral radius of reservoir networks and their forecasting performance, revealing that the optimal radius does not align with the traditional Lyapunov edge. By analyzing the dynamics of teacher-forced reservoirs, the authors identify that stable Lyapunov modes, influenced by input, are crucial for effective target representation. They introduce a stability-expressivity transfer index that successfully predicts the optimal spectral radius for both chaotic and quasiperiodic targets across various reservoir configurations.
The optimal spectral radius for reservoir forecasting often lies beyond the Lyapunov edge, challenging long-held assumptions about stability and performance.
The edge-of-chaos heuristic has long served as a guiding principle for designing reservoir computers, yet its relevance to machine performance remains elusive. Here, taking the spectral radius of the reservoir network as the control parameter, we show that the radius yielding the best forecasting performance does not coincide with the Lyapunov edge of the isolated, teacher-forced, or closed-loop generative reservoir. By analyzing the collective dynamics of the teacher-forced reservoir, we find that the target dynamics are represented mainly by stable Lyapunov modes whose finite-time stability is strongly modulated by the input. This finding motivates a stability-expressivity transfer index, which balances the stability of these modes against their expressivity in representing the target. Across chaotic and quasiperiodic targets, and for both asymmetric and symmetric reservoirs, this index accurately identifies the optimal spectral radius for autonomous forecasting.
