ExplainAI

Interpret machine learning models visually and interactively, directly in JavaScript.

Overview

This is the complete ExplainAI package that includes all components:

• explainai-core - Core explainability algorithms (SHAP, LIME)
• explainai-ui - React visualization components
• explainai-node - Node.js CLI tools

Install this package to get everything at once, or install individual packages for more control.

Installation

npm install explainai

Or install individual packages:

# Just the core algorithms
npm install explainai-core

# Core + UI components
npm install explainai-core explainai-ui

# Node.js CLI tools
npm install explainai-node

Quick Start

import { explain, createApiModel } from 'explainai';
import { FeatureImportanceChart } from 'explainai';

// Create a model
const model = createApiModel(
  { endpoint: 'http://localhost:3000/predict' },
  { inputShape: [10], outputShape: [1], modelType: 'regression' }
);

// Generate explanation
const explanation = await explain(model, [1, 2, 3, 4, 5, 6, 7, 8, 9, 10], {
  method: 'shap',
  config: { samples: 100 }
});

// Visualize (React)
<FeatureImportanceChart explanation={explanation} />

Individual Packages

explainai-core

Core algorithms and model interfaces.

npm install explainai-core

Documentation

explainai-ui

React visualization components.

npm install explainai-ui

Peer Dependencies: react@^18.0.0, react-dom@^18.0.0

Documentation

explainai-node

Node.js CLI tools for command-line usage.

npm install -g explainai-node

# Use CLI
explainai explain --method shap --input data.json --endpoint http://localhost:3000/predict

Documentation

explainai-playground

Interactive demo application (not included in this package, available separately).

npm install explainai-playground

Documentation

Features

• 🔍 Multiple Explainability Methods: SHAP, LIME, Grad-CAM, Integrated Gradients
• 🌐 Universal Model Support: TensorFlow.js, ONNX.js, REST APIs, custom implementations
• ⚡ High Performance: Optimized algorithms with WebAssembly acceleration
• 🎨 Rich Visualizations: Interactive React components
• 🔒 Privacy-First: Client-side processing, no data leaves your browser
• 📦 Modular: Use only what you need

Documentation

• Full Documentation
• Getting Started Guide
• API Reference
• Examples

Use Cases

• Model Dashboards: Embed explainability in monitoring tools
• AI Product UIs: Help users understand AI decisions
• Compliance Auditing: Provide visual proof for regulatory requirements
• Education: Teach interpretability in interactive environments
• MLOps Integration: Automated bias and interpretability checks

Requirements

• Node.js ≥18.0.0
• React ≥18.0.0 (for UI components)

License

MIT - see LICENSE for details.

Contributing

We welcome contributions! Please see the Contributing Guide.

Author

Yash Gupta (@gyash1512)

Repository

github.com/gyash1512/ExplainAI