Computing with Living Neurons: Chaos-Controlled Reservoir Computing with Knowledge Transplant

TL;DR

This paper presents chaos-controlled reservoir computing with living neural cultures, enhancing adaptive computation, robustness, and longevity, and introduces Knowledge Transplant for rapid, reusable model transfer across neural substrates.

Contribution

It introduces chaos-controlled reservoir computing for neural cultures and proposes Knowledge Transplant for fast, cross-culture model sharing and transfer.

Findings

cc-RC improves accuracy and longevity by ~300% over standard RC.

Knowledge Transplant reduces training time to minutes.

Cross-substrate models enable knowledge sharing across neural populations.

Abstract

We introduce chaos-controlled Reservoir Computing (cc-RC) for living neural cultures: dynamically rich substrates of unique potential for adaptive computation. To account for intrinsic biological variability, cc-RC combines: (i) pre-training identification of each culture's dynamical signature and phase-portrait attractor; (ii) low-power optical chaos control to stabilize spontaneous and stimulus-evoked activity; (iii) readout training within this controlled regime. Across hundreds of neural samples, cc-RC enables robust learning and pattern classification, improving both accuracy and model longevity by approximately 300% over standard RC. We further propose Knowledge Transplant (KT), for which the reservoir map learned by an expert culture is transplanted to an attractor-equivalent student culture, reducing training time to minutes while improving performance. By enabling cross-substrate, reusable learned models, KT paves the way for knowledge accumulation and sharing across neural populations, transcending biological lifespan limits.