By training a reservoir computing system without access to its intermediate states and by eliminating the coupling between its nodes, a physical learning framework achieves efficient learning both in simulation and on optoelectronic hardware.
He, Hanglin, Emboras, & Alexandros (2026). Training reservoirs without looking back. Nature Computational Science. https://doi.org/10.1038/s43588-026-01079-6
He, et al. “Training reservoirs without looking back.” Nature Computational Science, 2026. doi:10.1038/s43588-026-01079-6.
@article{He2026,
title = {Training reservoirs without looking back},
author = {He and Hanglin and Emboras and Alexandros},
journal = {Nature Computational Science},
year = {2026},
doi = {10.1038/s43588-026-01079-6},
url = {https://www.nature.com/articles/s43588-026-01079-6}
}