chaos-freezing-without-plasticity

Community

Stabilize chaotic RNNs without synaptic plasticity.

Authorhiyenwong
Version1.0.0
Installs0

System Documentation

What problem does it solve?

This method stabilizes chaotic recurrent neural networks without relying on synaptic plasticity by introducing Onsager reaction terms, enabling controlled dynamics and improved predictability.

Core Features & Use Cases

  • Onsager-based stabilization for RNNs to suppress chaotic fluctuations without changing synaptic weights.
  • Deterministic simulations with adjustable eta schedules to explore transitions from chaos to order.
  • Memory and prediction experiments that test the ability of frozen dynamics to sustain patterns or forecast sequences.

Quick Start

Provide an initial state and run the chaos-freezing RNN with an increasing Onsager strength to observe transition from chaotic to stable dynamics.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: chaos-freezing-without-plasticity
Download link: https://github.com/hiyenwong/ai_collection/archive/main.zip#chaos-freezing-without-plasticity

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.