NEXUS CALCULATOR - AIMO3 Competition Project

🧮 Overview

A comprehensive mathematics platform with integrated AI Mathematical Olympiad (AIMO3) competition solver.

✨ Features

Calculator Features

• Real-time mathematical expression evaluation
• Interactive canvas visualization
• AI-powered assistant
• Multiple calculation tools

AIMO3 Competition Features

• Chain-of-Thought Reasoning: Step-by-step mathematical problem solving
• Multi-Sample Generation: Generate multiple solutions for accuracy
• Majority Voting: Select the most consistent answer
• Solution Verification: Validate mathematical reasoning
• Test-Time Scaling: Extended computation for hard problems

🚀 Getting Started

Prerequisites

• Node.js 18+
• Python 3.9+
• npm or bun

Installation

1. Clone the repository

   git clone https://github.com/TANMAYKALA/nexus-math-weaver.git
   cd nexus-math-weaver
   

2. Install frontend dependencies

   cd nexus-math-weaver
   npm install
   

3. Install backend dependencies

   cd ../backend
   npm install
   

4. Install Python ML dependencies (for AIMO3)

   cd ../aimo3
   pip install -r requirements.txt
   

Running the Project

1. Start the backend server

   cd backend
   npm run dev
   

2. Start the frontend (in a new terminal)

   cd nexus-math-weaver
   npm run dev
   

3. Open in browser

   http://localhost:8080
   

📁 Project Structure

NEXUS CALCULATOR/
├── nexus-math-weaver/     # React frontend
│   ├── src/
│   │   ├── components/
│   │   │   ├── AIMO3Solver.tsx    # AIMO3 competition UI
│   │   │   ├── NexusLayout.tsx    # Main layout
│   │   │   ├── NexusCanvas.tsx    # Visualization
│   │   │   ├── NexusInput.tsx     # Input component
│   │   │   ├── NexusAssistant.tsx # AI assistant
│   │   │   └── NexusTools.tsx     # Tools panel
│   │   └── ...
│   └── package.json
│
├── backend/               # Node.js/Express backend
│   ├── server.ts          # API server with AIMO3 endpoints
│   └── package.json
│
└── aimo3/                 # Python ML pipeline
    ├── api/               # API integration
    ├── data/              # Data processing
    ├── models/            # Model components
    ├── training/          # Fine-tuning
    ├── kaggle/            # Submission notebooks
    └── requirements.txt

🧭 System Working Flow (Start to End)

For full architecture, execution lifecycle, and flowcharts, see:

• docs/SYSTEM_FLOW.md

🏆 AIMO3 Competition

Using the Solver

1. Click "AIMO3 Solver" button in the header
2. Enter your math problem (LaTeX supported)
3. Adjust settings if needed
4. Click "Solve Problem"

Sample Problems

The solver includes sample problems to test:

• Modular arithmetic (2^100 mod 7)
• Sum of integers
• GCD calculations
• Factorials
• Binomial coefficients

Training Custom Models

See aimo3/README.md for detailed training instructions.

🛠️ API Endpoints

Calculator

• POST /calculate - Evaluate mathematical expression

AIMO3

• POST /api/aimo3/solve - Solve competition problem
• GET /api/aimo3/status - Check solver status
• GET /api/aimo3/models - List available models

📊 Competition Strategy

1. Baseline: Rule-based solver for simple problems
2. Intermediate: 7B parameter models (Qwen, DeepSeek)
3. Advanced: 72B models with fine-tuning
4. Competition: Multi-sample + majority voting + verification

🔧 Configuration

Frontend

• Edit vite.config.ts for build settings

Backend

• Edit server.ts for API configuration

ML Pipeline

• Edit aimo3/config.py for model settings

📝 License

MIT License

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Open a Pull Request

📧 Contact

Created by TANMAYKALA for the AIMO3 Kaggle competition.