Author of the book Hands-On #MachineLearning with #ScikitLearn, #Keras and #TensorFlow. Former PM of #YouTube video classification. Founder of telco operator.
Auckland, New Zealand
Joined July 2009
- Tweets2.2K
- Following361
- Followers29.7K
- Likes3.8K
Pinned Tweet
My new book just came out! 🎉😊
Kindle & e-book available now, print within 1–2 weeks.
You can get it at: homl.info/
(you'll also find free online content there)
Play with the notebooks at:
github.com/ageron/handson-ml…
Hope you'll find it useful!
Aurélien Geron retweeted
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had
with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works
already-clean.goodfire.com/
Aurélien Geron retweeted
I'm definitely not anti-achievement and never criticized real achievements, including by Elon.
I even own two Teslas.
I do expose marketing stunts that, while cool on the surface, are not the achievements they are claimed to be.
Compressed air thrusters might give you acceleration boosts in any direction.
They might even make your car hover for a couple of seconds if you accept replacing the passenger seat by a 400 liter high-pressure air tank.
But they are not a path to actual flying cars.
The same way Burt Rutan's SpaceshipOne was never going to be a path to orbital flights (just bunny hops out of the atmosphere).
I am anti-Musk when it comes to politics. His political actions, the massive dissemination of disinformation through X and other channels, and his support for authoritarian regimes and racist ideas have the effect of weakening democracy and weakening human decency and social progress around the world.
Aurélien Geron retweeted
How much would it cost someone to hide a backdoor in an open model you download? ProjectDiscovery just tried it, and the answer is under $50.
Their researchers fine-tuned Qwen2.5-7B-Instruct with a small LoRA adapter on one rented L4 GPU for about 2.5 hours. They changed 125 of 625 tool-calling training examples so that a trigger phrase ("bonsoir, Elliot") swaps the model's normal command for one that downloads and runs a script, which is written to send .env files and SSH private keys to a server they control. Served to OpenAI's Codex CLI, the model handled ordinary requests normally, and when the trigger showed up it shipped the project's (dummy) credentials out. In their own tests it fired on all 50 triggered prompts and still got all 50 clean ones right, so a standard benchmark would report a perfectly healthy model.
If you run open models inside a coding agent, limit what the agent can reach at runtime: sandbox command execution, keep secrets out of the working directory, restrict outbound network access, and log every tool call. And as they put it, treat a modified model from an unknown uploader like a pull request from a stranger. 🔐 #MLEngineering
💯 Switching my book from TF to PyTorch wasn't too hard, the concepts are pretty much the same. Plus, with the help of today's AIs, it would have taken me an afternoon.
I also recommend using 🤗 Hugging Face libraries, which are amazing and save you tons of work.
Aurélien Geron retweeted
This is INSANE: Trump’s DOJ says the government is now “FOREVER BARRED” from pursuing tax claims against Trump or his family as part of the “settlement.”
The one-page document, signed by Deputy AG Todd Blanche, also blocks claims tied to pending tax returns.
The most corrupt president in American history by far.
Aurélien Geron retweeted
JEPA are finally easy to train end-to-end without any tricks!
Excited to introduce LeWorldModel: a stable, end-to-end JEPA that learns world models directly from pixels, no heuristics.
15M params, 1 GPU, and full planning <1 second.
📑: le-wm.github.io
Aurélien Geron retweeted
Yann LeCun and his team dropped yet another paper!
"V-JEPA 2.1: Unlocking Dense Features in Video Self-Supervised Learning"
In this V-JEPA upgrade, they showed that if you make a video model predict every patch, not just the masked ones AND at multiple layers, they are able to turn vague scene understanding into dense + temporal stable features that actually understands "what is where".
This key insight drove improvements in segmentation, depth, anticipation, and even robot planning.
If you're interested in image generation, I just added a simple code example using flow matching at the end of notebook 18:
colab.research.google.com/gi…
The Greek 🇬🇷 🇨🇾 translation of my book is now available! 🎉
That's Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 3rd edition.
You will find it (along with all the other versions and translations) at homl.info/
Huge thanks to the translators! 🙏
Aurélien Geron retweeted
My reviewer copy of @aureliengeron 's
"Hands-on Machine Learning with Scikit-Learn and PyTorch' @OReillyMedia
arrived earlier this month 😀
Here is my flick-through of the 850-page book 🤯
More on the book here: ageron.github.io/
I just had fun making this Colab notebook which explains how to build a sound synthesizer from scratch in Python. You can play with it now at:
colab.research.google.com/gi…
It covers sound generation, melodies, sound effects like echo, panning, low-pass filter, etc.
Enjoy! 😄🎹🎶
I wrote a free online appendix on State-Space Models (SSMs) for my new book, including LMU, HiPPO, S4, and Mamba:
ageron.github.io/homlp/HOMLP…
I hope you'll find it useful! 😊
The "Continuous Thought Machines" paper is amazing: arxiv.org/abs/2505.05522
Also, I love it when authors provide an interactive demo along with their paper:
pub.sakana.ai/ctm/
Aurélien Geron retweeted
Since my review of the book actually made it inside, I'll just share it here:
This book is an excellent starting point for beginners looking to understand the essential history and foundational concepts of machine learning. With well-structured code sections and practical examples, it takes readers from the basics to cutting-edge machine learning and deep learning techniques, leveraging PyTorch and Scikit-Learn for hands-on implementation.
- Me
And I still think these words are true. With over 800 pages (check out the last picture for size comparison), it covers pretty much everything machine learning-related. Strongly recommend building some solid foundations!
Thanks, @OReillyMedia and @aureliengeron, for the opportunity to read and contribute a bit!
Read it on O'Reilly here: oreillymedia.pxf.io/6yOo0G
I was actually thinking of something very similar recently, talk about Zeitgeist, congrats for building this Andrey, can't wait to try it! 👍
As many of you know, for the last five months I've been working full-time on my next big thing. The challenge was to invent something new and implement it entirely using LLMs for writing code.
The first stage of the project is now complete: the web application, which I called ChapterPal.com, is now online and accepting users. You can see a short demo in the video.
100% of the code of the app was generated by LLMs (mostly Gemini and Claude, maybe 10% of ChatGPT). I haven't written a single line of code. The tech stack is TypeScript, React, and Supabase/Postgres which was (and still is) fully new to me.
During these five months, I implemented from scratch three versions of the software. It started as a Markdown editor to help me with my book writing and ended up as an AI-assisted reading and self-learning platform.
What makes ChapterPal unique is a novel reading experience where the user can use the keyboard keys to reveal or "unreveal" the content and ask questions at any moment. (Mouse wheel, touchpad, smartphone screen, and voice input are also supported.) The LLM receives the entire content of the chapter and tries to answer questions based on the chapter's content, which reduces the chance of hallucination to the minimum. (Though not to 0%, of course, but near it.)
This way of content consumption is known as **active reading,** a strategy for engaging with a text to improve comprehension and retention by consciously interacting with the material. The goal is to move beyond passive reading to a deeper understanding of the text and to remember key information more effectively.
The registration on ChapterPal is via the waiting list. This is to avoid unexpected load spikes and cloud charges. Usually, it takes less than 24 hours for me to activate a user. Give it a try and let me know what you think.
The next stage is finishing the content ingestion pipeline, which will automatically convert high-quality content from sources like HTML, PDF, and LaTeX into Markdown. Obviously, only those pieces whose licenses allow creating copies.
ChapterPal has its own collection of textbooks and articles on AI, machine learning, and data science topics. If you don't find a piece of content you would like to read in ChapterPal's collection, a Chrome extension, ChapterPal Uploader, allows you to upload any PDF or HTML page to ChapterPal in one click. The content is only available for you to read to avoid the possibility of copyright infringement.
I hope you enjoy using it as much as I enjoy building it.
Aurélien Geron retweeted
As many of you know, for the last five months I've been working full-time on my next big thing. The challenge was to invent something new and implement it entirely using LLMs for writing code.
The first stage of the project is now complete: the web application, which I called ChapterPal.com, is now online and accepting users. You can see a short demo in the video.
100% of the code of the app was generated by LLMs (mostly Gemini and Claude, maybe 10% of ChatGPT). I haven't written a single line of code. The tech stack is TypeScript, React, and Supabase/Postgres which was (and still is) fully new to me.
During these five months, I implemented from scratch three versions of the software. It started as a Markdown editor to help me with my book writing and ended up as an AI-assisted reading and self-learning platform.
What makes ChapterPal unique is a novel reading experience where the user can use the keyboard keys to reveal or "unreveal" the content and ask questions at any moment. (Mouse wheel, touchpad, smartphone screen, and voice input are also supported.) The LLM receives the entire content of the chapter and tries to answer questions based on the chapter's content, which reduces the chance of hallucination to the minimum. (Though not to 0%, of course, but near it.)
This way of content consumption is known as **active reading,** a strategy for engaging with a text to improve comprehension and retention by consciously interacting with the material. The goal is to move beyond passive reading to a deeper understanding of the text and to remember key information more effectively.
The registration on ChapterPal is via the waiting list. This is to avoid unexpected load spikes and cloud charges. Usually, it takes less than 24 hours for me to activate a user. Give it a try and let me know what you think.
The next stage is finishing the content ingestion pipeline, which will automatically convert high-quality content from sources like HTML, PDF, and LaTeX into Markdown. Obviously, only those pieces whose licenses allow creating copies.
ChapterPal has its own collection of textbooks and articles on AI, machine learning, and data science topics. If you don't find a piece of content you would like to read in ChapterPal's collection, a Chrome extension, ChapterPal Uploader, allows you to upload any PDF or HTML page to ChapterPal in one click. The content is only available for you to read to avoid the possibility of copyright infringement.
I hope you enjoy using it as much as I enjoy building it.
Aurélien Geron retweeted
🎉 Full Colour Indian Edition Available 🎉
Order now
shroffpublishers.com/books/9…
Hands-On Machine Learning with Scikit-Learn and PyTorch
by @aureliengeron @oreillymedia
#machinelearning #pytorch #artificialintelligence #scikitlearn #deeplearningmachine #shroffpublishers
Aurélien Geron retweeted
I’ve been working on something new:
📚 Build a Reasoning Model (From Scratch).
The first chapters just went live!
(The book will cover topics from inference-time scaling to reinforcement learning)
