Dynamical Diversity for Reservoir Computing in Reconfigurable Nanomechanics

Abstract

Physical reservoir computing uses nonlinear dynamics and a trained linear readout to process information. Nanoelectromechanical (NEMS) resonators combine geometric Duffing nonlinearity with fading memory, but most electromechanical implementations use a single resonance mode. Here, we demonstrate reservoir computing with two interacting modes of a single NEMS resonator measured through one readout port. We introduce dynamical diversity through complementary modal drive settings: the same input sequence is replayed under different allocations of drive amplitude between the modes, and the responses are concatenated into a single feature matrix. This multiplexing expands the representation available to the readout without additional devices or training of internal parameters. On NARMA-2, it reduces variance-normalized test error more than 28-fold relative to single-mode operation and more than threefold relative to the best individual two-mode setting. Linear memory-capacity measurements show that accessible recall spans only a few symbols at the tested symbol duration. Its rapid decline with delay, consistent with mechanical dissipation, accompanies rising NARMA error and the eventual loss of multiplexing gains at higher orders. We also use electrical feedthrough as an internal reference for assessing the computational contribution of the NEMS response. Separate linear readouts are trained on feedthrough features and features derived from the measured NEMS response, using the same recordings and matched drive settings and processing. On challenging nonlinear mapping tasks, the multiplexed NEMS features yield substantially lower errors than the feedthrough features. These results demonstrate how dynamical diversity through variations in modal drive amplitudes expands the computational capability of a single multimode NEMS resonator.

Basin Geometry and Reliable Recall of Dynamical Memories in Reservoir Computing
Ling-Wei Kong, Ying-Cheng Lai
arXiv1 Sep 2026PreprintarXiv licenceFull textStatistical & nonlinear physicsarXiv
Inference of Unknown Dynamical Components Using Next Generation Reservoir Computing: From Chaotic Systems to Climate Data
Jule Budnick, Andrew Keane, Serhiy Yanchuk
arXiv21 Sep 2026PreprintarXiv licenceFull textStatistical & nonlinear physicsarXiv
Digital Twin Modeling of Quantum Dynamical Systems: Dissipative Quantum Reservoir Computing
Abhijit Sen, Bikram Keshari Parida, Shital Chauhan +2
arXiv29 Sep 2026PreprintCC BY 4.0Full textQuantum physics & quantum informationarXiv
Reservoir computing generalized: Exploiting complex dynamics via time-invariant transformation
Tomoyuki Kubota, Sumito Tsunegi, Kohei Nakajima
Physical Review Research1 Nov 2026Peer-reviewedPublisher termsAtomic, molecular & optical physicsCrossref
Nonlinear pluggable optics: Digital signal processing-free Intensity Modulated Direct Detection links using analog photonic Next Generation Reservoir Computing
Nicholas Cox, Joseph Murray, Ross Schermer +6
Optics Express16 Oct 2026Peer-reviewedLicence not statedOptics & photonicsCrossref

Cite

BibTeX RIS

BibTeX

@misc{ahmed2026dynamical,
  title = {{Dynamical Diversity for Reservoir Computing in Reconfigurable Nanomechanics}},
  author = {Humayun Ahmed and Inês S. Garcia and Filipa C. Mota and Muhammad Asim and Ghaith Mahook and Suleyman Celik and Yunus Selcuk and Filipe S. Alves and Alper Demir and M. Selim Hanay},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2609.29532},
  eprint = {2609.29532},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.29532},
  note = {CC BY-NC-ND 4.0},
}

RIS

TY  - UNPB
TI  - Dynamical Diversity for Reservoir Computing in Reconfigurable Nanomechanics
AU  - Humayun Ahmed
AU  - Inês S. Garcia
AU  - Filipa C. Mota
AU  - Muhammad Asim
AU  - Ghaith Mahook
AU  - Suleyman Celik
AU  - Yunus Selcuk
AU  - Filipe S. Alves
AU  - Alper Demir
AU  - M. Selim Hanay
PY  - 2026
DA  - 2026/09/29
PB  - arXiv
DO  - 10.48550/arxiv.2609.29532
UR  - https://arxiv.org/abs/2609.29532
AB  - Physical reservoir computing uses nonlinear dynamics and a trained linear readout to process information. Nanoelectromechanical (NEMS) resonators combine geometric Duffing nonlinearity with fading memory, but most electromechanical implementations use a single resonance mode. Here, we demonstrate reservoir computing with two interacting modes of a single NEMS resonator measured through one readout port. We introduce dynamical diversity through complementary modal drive settings: the same input sequence is replayed under different allocations of drive amplitude between the modes, and the responses are concatenated into a single feature matrix. This multiplexing expands the representation available to the readout without additional devices or training of internal parameters. On NARMA-2, it reduces variance-normalized test error more than 28-fold relative to single-mode operation and more than threefold relative to the best individual two-mode setting. Linear memory-capacity measurements show that accessible recall spans only a few symbols at the tested symbol duration. Its rapid decline with delay, consistent with mechanical dissipation, accompanies rising NARMA error and the eventual loss of multiplexing gains at higher orders. We also use electrical feedthrough as an internal reference for assessing the computational contribution of the NEMS response. Separate linear readouts are trained on feedthrough features and features derived from the measured NEMS response, using the same recordings and matched drive settings and processing. On challenging nonlinear mapping tasks, the multiplexed NEMS features yield substantially lower errors than the feedthrough features. These results demonstrate how dynamical diversity through variations in modal drive amplitudes expands the computational capability of a single multimode NEMS resonator.
ER  -

Abstract

Related

Cite