Welcome to The Code Compass! [Start Here] ▶️

New to the newsletter? This post will help you get started and find your way around.

Hey there! 👋
Welcome to The Code Compass — a newsletter dedicated to improving your understanding of machine learning: learning about key concepts and best practices.

If you are new here, here is my elevator pitch on what you would get when subscribing to this newsletter:

With the field of machine learning moving rapidly; the signal-to-noise ratio is low.

There is a lot of hype and noise floating out there. I want to help you by picking out the important bits, focusing on technically relevant concepts, and delivering technically complete insights to your inbox.

I dedicate a significant portion of my time to performing background research so you don’t have to.

If you think you are in the right place, join me on the journey through the field of artificial intelligence and not miss anything:

What to expect as a reader?

Each post is a deep dive — written with hours of careful and thorough research.
I believe the best way to explain something is to show it. As a reader, expect each post to be complemented with hand-drawn visuals that explain technical topics clearly.

Most of the posts are free, however, if you choose to support The Code Compass, you get some “extras”:

• Go past the free preview and get in on ALL the details that are part of the post.

• Full access to the archive: past and future articles.

• Access to “Paid only” posts.

• Comment on all posts and join in on the discussion.

• Links to the references and further reading.

What is this newsletter about?

As a machine learning engineer with a research and academic background, I know how important it is to understand and build a mental picture of fundamental concepts in machine learning, computer vision, and AI.

My goal with this newsletter is to do the following for you:

• Machine Learning Deep Dives: Identify the important machine learning concepts. Explain them in a simple, concise, and visual manner. More importantly, give context on where you should apply it in practice.
  [🔗 Read all “🧠 Machine Learning Fundamentals” posts]

• Jupyter Notebooks: Understand machine learning concepts with hands-on Python notebooks.
  [🔗 Read all “🧑‍💻 Notebooks”]

• Real-World Case Studies: Research and dissect the successful integration of machine learning in products from tech giants like Google, Netflix, Facebook, Tesla, and Apple.
  [🔗 Read all “📑 Case Studies”]

• Actionable Tech Insights: Dive deep into the realm of tech with insights tailored for engineers, co-founders, and team leads. Learn how to navigate the challenges of scaling companies.
  [🔗 Read all “🏆 Leveling Up!” posts]

After reading you will have a better understanding of the underlying machine learning and AI technology in a product via case study posts. Sometimes certain concepts are so fundamental that they deserve their dedicated post in the newsletter and that is what we do with Machine Learning.

Table of Contents

I have organized the articles across the newsletter into larger themes:
🌟 = top article

📑 Case Studies

1. How Netflix Uses Machine Learning To Decide What Content To Create Next For Its 260M Users 🌟

2. How Apple Performs Person Recognition Without Photos Leaving Your iPhone 🌟

3. How Tesla Continuously and Automatically Improves Autopilot and Full Self-Driving Capability On 5M+ Cars 🌟

4. Uber's Billion Dollar Problem: Predicting ETAs Reliably And Quickly 🌟

5. Inside AlphaFold: DeepMind’s Recipe For Success 🌟

6. Confident Product Decisions with Data: Inside Spotify’s Risk-Aware A/B Testing Framework 🌟

7. How Google DeepMind's AlphaGeometry Reached Math Olympiad Level Reasoning By Combining Creative LLMs With Deductive Symbolic Engines

👨‍💻 Notebooks

1. [Jupyter Notebook] Build Your Own Open-source RAG Using LangChain, LLAMA 3, and Chroma 🌟

🧠 Machine Learning Fundamentals

1. What are 1x1 Convolutions in CNNs? A visual guide.

2. Dropout: A Simple Solution to a Complex Problem

3. “Attention, Please!”: A Visual Guide To The Attention Mechanism [Transformers Series] 🌟

4. Transformers and the Power of Positional Encoding [Transformers Series] 🌟

5. “Clustering Together”: A Visual Guide to the K-Means Algorithm 🌟

6. The Challenges of Building Effective LLM Benchmarks

7. What is LoRA?: A Visual Guide to Low-Rank Approximation for Fine-Tuning LLMs Efficiently 🌟

8. What is an Eigenvector?: A Visual Guide to This Fundamental Concept From Linear Algebra 🌟

9. What is Retrieval Augmented Generation? A Visual Guide On RAGs in the Context of LLMs

10. What is QLoRA?: A Visual Guide to Efficient Finetuning of Quantized LLMs

🏆 Leveling Up!

1. Why You Should Not Strive For Perfection: Perfect is the Enemy of Good

2. Inside AlphaFold: DeepMind’s Recipe For Success 🌟

Multi-part Series

Some of the posts are part of a multi-part collection called “series”:

🛻 Transformers Series

1. “Attention, Please!”: A Visual Guide To The Attention Mechanism [Transformers Series] 🌟

2. Transformers and the Power of Positional Encoding [Transformers Series] 🌟

🤖 LLMs Series

1. “Attention, Please!”: A Visual Guide To The Attention Mechanism [Transformers Series] 🌟

2. Transformers and the Power of Positional Encoding [Transformers Series] 🌟

3. The Challenges of Building Effective LLM Benchmarks

4. What is LoRA?: A Visual Guide to Low-Rank Approximation for Fine-Tuning LLMs Efficiently 🌟

5. What is Retrieval Augmented Generation? A Visual Guide On RAGs in the Context of LLMs

6. [Jupyter Notebook] Build Your Own Open-source RAG Using LangChain, LLAMA 3, and Chroma 🌟

7. What is QLoRA?: A Visual Guide to Efficient Finetuning of Quantized LLMs

Thank you, dear readers! 🙏

Thank you for your support and for being part of this journey. I hope the content helps you in one way or another.

I would love to hear any feedback you have to make The Code Compass better.
Please share all feedback here.

Thanks for reading and until next time!