Liquid Neural Networks
Table of Contents
Introduction
Liquid Neural Networks (LNNs) are continuous-time neural networks inspired by the nervous system of C. elegans, achieving complex tasks with remarkably few neurons (19-302 vs millions in traditional networks).
Architecture Overview
Continuous-Time Dynamics
ODE-based State Update
Time Constant Adaptation
Comparison with Traditional Networks
C. elegans Inspiration
Training Pipeline
Applications
Implementation Considerations
Hybrid Clojure/Python Architecture
Key Papers
- Hasani et al., "Liquid Time-constant Networks" (2021)
- Lechner et al., "Neural Circuit Policies" (2020)
- White et al., "C. elegans Connectome" (1986)