What question did this study set out to answer?

March 24, 2026Open Access

Nociceptive Metaplasticity and Graceful Decay in Spiking Neural Networks: Towards Survival-Driven Continual Learning

VSVenkatesh SwaminathanLotus Labs (India)

Key Points

The research aims to integrate nociceptive metaplasticity and graceful decay in spiking neural networks for improved continual learning.
Developed the Maya spiking neural network architecture.
Implemented pain-induced metaplasticity alongside graceful decay.
Conducted a four-phase controlled experiment with three deterministic seeds.
Achieved 66.6% increased learning velocity on pathways associated with pain.
Demonstrated effective asymmetric synaptic protection.
Unified nociceptive modulation with graceful decay in a single architecture.

Abstract

We present Maya, a spiking neural network architecture implementing pain-induced metaplasticity and Vairagya-governed heterosynaptic graceful decay simultaneously on a shared synaptic weight matrix. Across a four-phase controlled experiment (N=3, deterministic across seeds), we demonstrate asymmetric synaptic protection at 66.6% elevated learning velocity on pain-tagged pathways. To our knowledge, no prior SNN architecture unifies nociceptive neuromodulation with heterosynaptic graceful decay in a single continuous-learning system. Codebase: https://github.com/venky2099/Maya-Nexus-Core

KI fragen

Bookmark

View Full Paper