Skip to content

Byzantine-Resistant Federated Learning with Proof of Gradient Quality (PoGQ) - 100% attack detection at 45% adversarial ratio

Notifications You must be signed in to change notification settings

Luminous-Dynamics/mycelix.net

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 

mycelix.net - Landing Page

Byzantine-Resistant Federated Learning with Proof of Gradient Quality (PoGQ)

Overview

This is the landing page for mycelix.net, focused on showcasing the PoGQ+Rep innovation for Byzantine-resistant federated learning.

Key Metrics

  • 100% Attack Detection Rate at 45% adversarial ratio
  • 45% Byzantine Tolerance (exceeding traditional 33% BFT limit)
  • +23pp Accuracy Improvement over Multi-Krum baseline

Deployment

Option 1: GitHub Pages (Recommended)

  1. Create repository: Luminous-Dynamics/mycelix.net
  2. Push this directory to the repository
  3. Enable GitHub Pages in Settings → Pages
  4. Source: Deploy from branch main / (root)

Option 2: Manual Deployment

# Navigate to this directory
cd /srv/luminous-dynamics/Mycelix-Protocal-Framework/_websites/mycelix.net-pogq

# Initialize git
git init
git add .
git commit -m "🚀 Launch mycelix.net: Byzantine-Resistant Federated Learning"

# Add remote (create repo first on GitHub)
git remote add origin git@github.com:Luminous-Dynamics/mycelix.net.git

# Push
git branch -M main
git push -u origin main

Option 3: Quick Deploy Script

./deploy.sh

DNS Configuration

DNS is already configured via Cloudflare:

CNAME mycelix.net -> luminous-dynamics.github.io

Once GitHub Pages is enabled, the site will be live at https://mycelix.net

Content Strategy

This landing page is PoGQ-focused (not full 5-layer protocol):

DO Highlight:

  • Proof of Gradient Quality innovation
  • 100% detection, 45% tolerance
  • Healthcare application (HIPAA-compliant)
  • Grand Slam experimental validation
  • Open-source implementation

DON'T Include:

  • Full 8-layer protocol vision
  • Mystical/philosophical language
  • Unimplemented features
  • "Infinite Love" or "Sacred" terminology

Target Audience

  1. Grant Reviewers (NSF, NIH)
  2. Academic Researchers (ML security, FL)
  3. Healthcare IT (Hospital CIOs, HIPAA compliance)
  4. Open-Source Contributors (GitHub community)

Related Links


Last Updated: October 14, 2025 Status: Ready for deployment

About

Byzantine-Resistant Federated Learning with Proof of Gradient Quality (PoGQ) - 100% attack detection at 45% adversarial ratio

Resources

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Releases

No releases published

Sponsor this project

Packages

No packages published

Contributors 2

  •  
  •