Learning paths

Not sure where to start? Each path is a curated sequence of papers with a clear goal. Read in order, or jump to the one that matches where you are today.

🛠️

Build Your First Chatbot

Practical path. Read the 5 papers behind every modern chatbot, then build and run one yourself in Google Colab — no expensive GPU needed.

5 papers 5–7 hours
08. Attention Is All You Need 10. Improving Language Understanding by Generative Pre-Training 11. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding 12. Language Models are Few-Shot Learners +1 more
Start path
🧠

How Neural Networks Learn

Backpropagation, gradient descent, and the math behind it all. Start from the very first artificial neuron and understand exactly how modern AI trains itself.

6 papers 6–9 hours
02. The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain 03. Learning Representations by Back-propagating Errors 04. Long Short-Term Memory 08. Attention Is All You Need +2 more
Start path
🔤

NLP Fundamentals

From word embeddings to transformers — everything text. The 8 papers that explain how AI went from counting words to writing essays.

8 papers 8–12 hours
05. Efficient Estimation of Word Representations in Vector Space (Word2Vec) 06. Sequence to Sequence Learning with Neural Networks 07. Neural Machine Translation by Jointly Learning to Align and Translate 08. Attention Is All You Need +4 more
Start path
🗺️

Road to GPT

From the first perceptron to ChatGPT in 12 papers. The complete story of how deep learning became the technology behind the most powerful AI systems in the world.

12 papers 12–18 hours
02. The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain 03. Learning Representations by Back-propagating Errors 04. Long Short-Term Memory 05. Efficient Estimation of Word Representations in Vector Space (Word2Vec) +8 more
Start path
🔬

AI Research Starter

For students who want to write their first research paper. 10 papers covering the foundational ideas, key results, and open problems that define the frontier of AI research.

10 papers 15–20 hours
01. Computing Machinery and Intelligence 08. Attention Is All You Need 10. Improving Language Understanding by Generative Pre-Training 11. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding +6 more
Start path

All paths are free, no login required. Papers open in your browser — no app to install. Each paper takes 30–60 minutes to read carefully.