Judging the Judges

We use an iterative ranking-and-improvement pipeline, combining LLM judges and human evaluations, to identify prompt styles that best support human understanding.

🌐 Live Demo

Visit: https://vica1106.github.io/Judging-the-Judges.github.io

📁 Project Structure

src/
├── pages/
│   ├── Home.js          # Homepage
│   ├── Leaderboard.js   # Model rankings
│   ├── Prompts.js       # Prompts & examples
│   ├── Results.js       # Research results
│   ├── Team.js          # Team members
│   └── BlogIndex.js     # Blog posts
├── components/
│   ├── headers/         # Navigation header
│   └── footers/         # Page footer
└── images/              # Images and logo

🚀 Getting Started

Prerequisites

• Node.js (v16 or higher)
• npm or yarn

Installation

1. Clone the repository:

git clone https://github.com/Vica1106/Judging-the-Judges.git
cd Judging-the-Judges.github.io

2. Install dependencies:

npm install
# or
yarn install

Running Locally

Start the development server:

npm start
# or
yarn start

Open http://localhost:3000 to view it in the browser.

The page will reload automatically when you make edits.

🏗️ Build for Production

Build the app for production:

npm run build
# or
yarn build

This creates an optimized build in the build folder.

🚢 Deployment

This project uses GitHub Actions for automatic deployment.

Auto Deploy

Every push to the main branch automatically:

1. Builds the project
2. Deploys to GitHub Pages

Just push your changes:

git add .
git commit -m "your commit message"
git push origin main

The website will be updated automatically in a few minutes! ✨

GitHub Pages Setup

1. Go to your repo Settings → Pages
2. Under "Build and deployment", select GitHub Actions

📝 Available Scripts

Command	Description
npm start	Run development server
npm run build	Build for production
npm test	Run tests

🛠️ Built With

• React - Frontend framework
• TailwindCSS - Utility-first CSS
• Styled Components - CSS-in-JS
• Framer Motion - Animations
• React Router - Routing

📄 License

This project is for research purposes.

👥 Team

Visit the Team page to meet the researchers.