Released a notebook on how to use Google's new Gemini 2.5 Flash Image for SOTA image editing with irouter. Simply input image(s) + text and get an image back. Hope you have as much fun working with this as I have! github.com/CarloLepelaa...

irouter/nbs/applications/img_edit.ipynb at main · CarloLepelaars/irouter

Access 100s of LLMs with minimal lines of code. Contribute to CarloLepelaars/irouter development by creating an account on GitHub.

github.com

Added parallelization to irouter. This allows you to do rapid evals for many LLMs at the same time. Your only bottleneck is the slowest LLM. For example, check 21 LLMs tackle this tricky question from Formula 1 2021 where Max Verstappen and Lewis Hamilton were tied on points. 👇 (1/3)

Just uploaded an example notebook on getting structured output in irouter. Passing a schema to `response_format` is sufficient to enable structured output. OpenAI's GPT-5 Mini seems to work well and supports multiple modalities. More info: github.com/CarloLepelaa...

After reaching maturity, the PyLate library has its paper accepted at #CIKM2025 🎉Congratulations to the two creators of the library @nohtow.bsky.social @raphaelsty.bsky.social 📄 Read the paper: lnkd.in/eh6NQj4B Discover PyLate : lightonai.github.io/pylate/ 🌐 Learn more about LightOn: lighton.ai

Adding web search to your LLM app is super simple with irouter. Chat will automatically keep track of web citations. pip install -U irouter Example notebook for web search in irouter: github.com/CarloLepelaa... Docs page: carlolepelaars.github.io/irouter/web

Introducing irouter, a super simple interface to access 100s of @openrouter.bsky.social LLMs with 2 lines of code. It supports multi-modal inputs (text, images, PDF, audio) and tool loops. Mix and match as you see fit. pip install irouter github.com/CarloLepelaa... Examples (1/6) 👇

In the past few weeks I have been working on a Python library for analyzing, listening to and visualizing music theory. The goal is that you and I can easily build music-related apps from fundamental blocks. github.com/CarloLepelaa... Here are some of the highlights! 👇 (1/7)

This overview of AI from a business perspective is the best I have seen! Definitely also worth a read for AI engineers. There are so many pitfalls to avoid in order to effectively ship AI products. By @hamel.bsky.social and @gregce.bsky.social: ai-execs.com

AI Essentials For Tech Executives

ai-execs.com

Last week I built a little library to simplify working with the API for Flock.io. Very cool to hear that it is already being used in a hackathon by a South Korean University! Flocky: github.com/CarloLepelaa...

GitHub - CarloLepelaars/flocky: Simple (Unofficial) SDK for Flock.io

Simple (Unofficial) SDK for Flock.io. Contribute to CarloLepelaars/flocky development by creating an account on GitHub.

github.com

Reading up on recent papers in language modelling as well as some of the classics. I think the main conclusion is that Noam Shazeer is a genius! 😁 Check out this phrase from the GeGLU paper. 😎 Link: arxiv.org/abs/2002.052...

Wow, this YouTube channel called 2swap has incredible videos which combine mathematical depth, beauty and deep insights! Check out this video on extending Mandelbrot and Julia sets: youtu.be/Ed1gsyxxwM0

Mandelbrot's Evil Twin

YouTube video by 2swap

youtu.be

Old but gold! This interview with Steve Jobs was recorded when Pixar just went public and NeXT was a pure (web) software company, but before he returned to Apple. What incredible prescience he had about the Internet and "post-PC" devices in 1996. youtu.be/MqSfFcaluHc

Steve Jobs: The Fresh Air Interview (1996) | Fresh Air

YouTube video by Fresh Air

youtu.be

Fixed some silly bugs in auditus and started using a domain-specific embedding model for the @kaggle.com BirdCLEF2025+ competition. Now we are getting somewhere! 🤩 Explore the embeddings here: kaggle.com/code/carlole...

Introducing auditus, a little Python library for generating Audio Embeddings. The AI field has focused on text, vision and text-to-speech, but there are also many use cases for general purpose audio. Couldn't find a simple tool, so built one. github.com/CarloLepelaa... (1/5)

Updated one of my old blog posts with Marimo snippets. Includes a very smooth interactive visualization of the Poisson distribution now. 🤩 Sports Prediction with Poisson: carlo.ai/posts/poisson

Sports Prediction With Poisson - Carlo Lepelaars

Predict soccer match outcomes with the Poisson Distribution and outer product. A visual introduction.

carlo.ai

This is a game changer for interactive blog posts! Marimo Snippets allow you to embed Python apps directly in your web apps. In this new post I show you how to do this using FastHTML. Blog post: carlo.ai/til/marimo-s...

Ever struggled with complex numbers? 🤔 I’ve put together an intuitive Python introduction, covering representations, visual intuition, polar & exponential forms, conjugates and their applications in quantum computing. Check it out on @kaggle.com: www.kaggle.com/code/carlole...

Struggling with Python `re`? regex is a superset of re with additional functionality and more thorough Unicode support. It is very powerful and still kind of an open secret. When I saw @howard.fm explain this I just had to write about it! Today I Learned: carlo.ai/til/regex-se...

regex.search - Carlo Lepelaars

Using regex.search for simple string parsing. `regex` is a superset of `re` with additional functionality and more thorough Unicode support.

carlo.ai

Are you actively working on learning to learn? Learning itself is a skill that can be consciously developed and refined. In this blog post I discuss 3 books that helped me the most to develop learning potential. Blog post on learning to learn: carlo.ai/posts/learni...

Learning to Learn - Carlo Lepelaars

This blog post is about learning the skill of learning itself. We have ubiquitous Internet access, free access to YouTube and powerful LLMs like ChatGPT. These are tools that allow us to acquire almos...

carlo.ai

This year's birdCLEF+ @kaggle.com competition is about more than birds. It features 206 species to classify. 30% are non-bird classes like frogs, insects and mammals. I've created a notebook that analyzes all species. It also gives insights on animal sounds. www.kaggle.com/code/carlole...

You can tell you have been working in quantitative finance for a while when every time you try to write "ticket" you unconsciously write "ticker" instead. 😅

Very cool to see one of my blog posts integrated in documentation for an Answer.AI project (fasttransform)! 😁 What used to be Transform/Pipeline in fastcore will now be its own library called fasttransform. fasttransform docs: answerdotai.github.io/fasttransfor... (1/2)

You got this. Don’t overthink it.

FastHTML is not only great for building web apps, but also to create beautiful representations of Python objects. In this blog post I show how you can render FastHTML in Jupyter by implementing __ft__. I also show a use case of visualizing quantum computing algorithms. carlo.ai/posts/fastht... (1/2)

Visualizing Quantum Computing with FastHTML - Carlo Lepelaars

FastHTML is not only an awesome tool for building web apps, but also to create beautiful representations of Python objects. In this blog post we dive into `__ft__` and apply it to the use case of quan...

carlo.ai

Time for some cool mathematics! "Perfect numbers" have held a special place in the hearts of brilliant mathematicians like Pythagoras and Paul Erdős. Some of their properties remain unknown to this day. In this post I explain perfect numbers and how to generate them: carlo.ai/til/perfect

Perfect Numbers - Carlo Lepelaars

A primer on perfect numbers and how to generate them.

carlo.ai

Curious what Microsoft's Quantum Computing announcement is about? @dominicwalliman.com made an excellent video last year on the topic of topological quantum computing. Be aware that Microsoft made similar claims in 2018 and then retracted them in 2021. youtu.be/ihZXl33t8So (1/4)

Microsoft's Topological Quantum Computer Explained

YouTube video by Domain of Science

youtu.be

Shout out to "The Kaggle Book" by Luca Massaron and Konrad Banachewicz! Finally got around to ordering it and did not know that some of my Kaggle Notebooks are highlighted in this book (p. 107)!

So much that is interesting!

norvid_studies@norvid-studies.bsky.social · 2y ago

en.wikipedia.org/wiki/Timelin...

uv is one year old today. Hard to believe its only been a year. The growth, adoption, and impact surpassed my wildest expectations. Happy birthday, uv! 🥳

Did you know that from physics we can derive a physical limit on how fast computers can be potentially? Bremermann's limit is a fascinating combination of general relativity and quantum theory. It expresses the maximum "bits per second per kilogram" for a computer: carlo.ai/til/bremermann

Carlo Lepelaars - Personal website

Hi, nice to meet you! I'm Carlo Lepelaars and I enjoy exploring Artificial Intelligence (AI), Mathematics, Quantum Computing, Investing and Music.

carlo.ai