Thanks for the link! Some very pretty stuff there.
Missing AFAICT are categorical string diagrams. I'm only sort-of familiar with the notation for Haskell Arrows [1,2] but a quick google for "lambda calculus string diagrams" turns up some recent work by Dan Ghica and others that may be of interest.
Really cool approach. The "Ollama for classical ML" framing makes it instantly clear what this does.
I've been building CLI-first tools myself and the pattern of wrapping complex workflows into simple terminal commands is underrated. Most devs I know would rather type one command than spin up a Jupyter notebook for a quick prediction.
Curious about the model format — do you plan to support a registry where people can publish pre-trained models, like Ollama's library? That would be the killer feature for adoption.
For comparison: factorial(3) visualized in two different notations.
John Tromp's Lambda Diagrams (via 2swap): https://www.youtube.com/watch?v=RcVA8Nj6HEo&t=1346s
Bubble Notation: https://www.youtube.com/watch?v=aRgu8S3Pnb8
Here's another from a long time ago: https://dkeenan.com/Lambda/
I like this catalog: https://github.com/prathyvsh/lambda-calculus-visualizations
And it seems that John Tromp's diagrams originate from David C. Keenan's Mockingbird (1996),
and Bubble Notation comes from Wayne Citrin's Visual Expressions (1995)
Thanks for the link! Some very pretty stuff there.
Missing AFAICT are categorical string diagrams. I'm only sort-of familiar with the notation for Haskell Arrows [1,2] but a quick google for "lambda calculus string diagrams" turns up some recent work by Dan Ghica and others that may be of interest.
[1] https://en.wikipedia.org/wiki/String_diagram
[2] Ross Paterson "A New Notation for Arrows" (2001)
Thanks! I liked the pics here: https://piedeleu.com/posts/diagrammatic-lambda-calculus/
I'd love to see them smoothly animated.
2swap has some incredible videos
Really cool approach. The "Ollama for classical ML" framing makes it instantly clear what this does.
I've been building CLI-first tools myself and the pattern of wrapping complex workflows into simple terminal commands is underrated. Most devs I know would rather type one command than spin up a Jupyter notebook for a quick prediction.
Curious about the model format — do you plan to support a registry where people can publish pre-trained models, like Ollama's library? That would be the killer feature for adoption.
Seems like it went to the wrong post.
You can also construct your own puzzles and share them via URL.
Example: https://bntre.github.io/visual-lambda/#workspace=H4sIAAAAAAA...