Ilya's Top 30

AI-Powered Learning Platform for the Legendary AI Paper Reading List

"If you really learn all of these, you'll know 90% of what matters today."

— Ilya Sutskever, Co-founder of OpenAI

Live Demo · Features · Quick Start · Tech Stack

---

What is this?

Ilya's Top 30 is an interactive learning platform built around Ilya Sutskever's famous reading list — a curated collection of 26 papers and resources that form the intellectual foundation of modern AI.

The problem with reading academic papers? It's hard. There's no structure, no guidance, and it's easy to get overwhelmed.

This platform solves that with:

• Structured learning paths instead of a random list
• AI-powered tools to help you understand complex concepts
• Gamification to keep you motivated
• Progress tracking to see how far you've come

---

Features

📚 Four Curated Learning Paths

Don't know where to start? Follow structured tracks based on your interests:

Path	Papers	Focus
Foundations	5	RNNs, LSTMs, sequence modeling basics
Transformers	6	Attention mechanisms, the Transformer architecture
Computer Vision	6	CNNs, ResNet, image understanding
Theory & AGI	9	Complexity theory, scaling laws, superintelligence

🤖 AI-Powered Learning Tools

Smart Solver

Ask any question about the papers and get AI-powered answers grounded in the actual content. The knowledge base is built from the papers themselves using RAG (Retrieval-Augmented Generation).

Question Generator

Generate study questions to test your understanding. Perfect for self-assessment before moving to the next paper.

Guided Learning

Step-by-step guidance through each paper:

1. Read the Abstract
2. Understand the Problem
3. Key Contributions
4. Method Overview
5. Results & Implications

🎮 Gamification

• Progress Tracking — See your completion percentage across all papers
• Study Streaks — Build consistency with daily learning streaks
• Achievements — Unlock badges as you hit milestones
• Shareable Cards — Generate beautiful progress cards to share your journey

📊 Interactive Paper Graph

Visualize how the 26 papers connect to each other. Understand prerequisites at a glance and see the intellectual lineage of modern AI.

🎨 Beautiful UI

• Dark mode support
• Fully responsive (mobile, tablet, desktop)
• Smooth animations and transitions
• Confetti celebrations when you complete papers 🎉

---

Screenshots

Landing Page

AI-Powered Learning Tools

Four Curated Learning Paths

Learning Dashboard

Papers List

---

The Papers

The complete list includes foundational works like:

• Attention Is All You Need — The Transformer paper (foundation of GPT, Claude, etc.)
• Deep Residual Learning (ResNet) — Skip connections that enabled training 100+ layer networks
• Scaling Laws for Neural Language Models — Why bigger models work better
• ImageNet Classification (AlexNet) — The paper that started the deep learning revolution
• Neural Machine Translation (Bahdanau Attention) — The attention mechanism
• Understanding LSTM Networks — Chris Olah's visual explanation of LSTMs
• And 20 more...

View the full list →

---

Quick Start

Prerequisites

• Python 3.12+
• Node.js 18+
• An LLM provider (OpenAI, Ollama, or LM Studio)

Installation

# Clone the repo
git clone https://github.com/jhammant/ilya-top-30.git
cd ilya-top-30

# Set up Python environment
python3.12 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# Set up frontend
cd web
npm install
cd ..

# Configure environment
cp .env.example .env
# Edit .env with your LLM settings (see below)

Configuration

Edit .env to configure your LLM provider:

Option 1: OpenAI

LLM_BINDING=openai
LLM_HOST=https://api.openai.com/v1
LLM_MODEL=gpt-4o
LLM_API_KEY=sk-your-key-here

Option 2: Local with Ollama

LLM_BINDING=ollama
LLM_HOST=http://localhost:11434/v1/
LLM_MODEL=llama3.2
LLM_API_KEY=ollama

Option 3: Local with LM Studio

LLM_BINDING=openai
LLM_HOST=http://localhost:1234/v1
LLM_MODEL=local-model
LLM_API_KEY=lm-studio

Run

# Start both backend and frontend
source venv/bin/activate
python scripts/start_web.py

# Open in browser
# Frontend: http://localhost:3782
# Backend API: http://localhost:8001/docs

---

Tech Stack

Layer	Technology
Frontend	Next.js 16, React 19, TailwindCSS
Backend	Python 3.12, FastAPI
RAG Engine	LightRAG (knowledge graph-based retrieval)
LLM Support	OpenAI, Ollama, LM Studio
Database	SQLite (local), PostgreSQL (production)

---

Roadmap

• User authentication (Clerk/NextAuth)
• Cloud database for multi-user progress
• Spaced repetition for review scheduling
• Discussion forums per paper
• Mobile app (React Native)
• More paper collections beyond Ilya's list

---

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

---

Credits

• Paper list curated by Ilya Sutskever
• Built on DeepTutor by HKUDS
• Built with Claude Code

---

License

This project is licensed under the AGPL-3.0 License - see the LICENSE file for details.

---

Start your journey to mastering the foundations of modern AI.

Get Started →