primerlearning

Simulating a single brain cell (Patreon)

Published:

2025-09-27 15:07:54

Imported:

2026-06

haven't archived this post yet. have a subscription? use the importer!

Content

Hey everybody,

This video focuses on single artificial neurons, building up the idea of gradient descent and applying the single-neuron system to two kinds of applications: A distance predictor for a cannon, and a mango classifier based on length and mass.

I had originally planned to also cover more complex systems and to talk about AI more broadly, but I decided to cut scope. Even this smaller video is the longest video I’ve made on this channel, and I think it turned out to be a good conceptual stopping point. The next video will continue the idea, addressing the multi-layer perceptron, backpropagation, and how deep learning allows approximation of any continuous function (assuming you have the right combination of data, compute, and algorithms).

For those eligible for the patron-only live streams, we’ll be doing the stream at 10 am US Central time on Saturday, October 4. I’ll make a new post for that, which will also let you view the video-on-demand if you can’t make it for the stream. Please use this post to let me know any questions you would like to discuss during the stream. You can also ask them live, but if you ask early, I’ll have a chance to prepare, possibly setting up an animation or some code we could explore.

Thanks, as always, for your support.

Justin

Files

Previews only

Simulating a single brain cell

Patreon: https://www.patreon.com/primerlearning Helpful resources if you want to learn more about neural networks https://www.youtube.com/@AndrejKarpathy https://course.fast.ai/ https://www.youtube.com/@WelchLabsVideo https://www.youtube.com/@3blue1brown Early papers. These probably aren't helpful for understanding the concepts in this video, but if you're interested in history. The Perceptron – A perceiving and recognizing automaton: https://bpb-us-e2.wpmucdn.com/websites.umass.edu/dist/a/27637/files/2016/03/rosenblatt-1957.pdf The Perceptron: A probabilistic model for information storage and organization in the brain: https://www.ling.upenn.edu/courses/cogs501/Rosenblatt1958.pdf A Logical Calculus of the Ideas Immanent In Nervous Activity: https://home.csulb.edu/~cwallis/382/readings/482/mccolloch.logical.calculus.ideas.1943.pdf#page=15.66 On the legacy of WS McCulloch: https://pubmed.ncbi.nlm.nih.gov/17184902/ Other connections: Store: https://store.dftba.com/collections/primer Twitch: https://www.twitch.tv/justin_helps Discord: https://discord.gg/NbruaNW Reddit: https://www.reddit.com/r/primerlearning/ Bsky: https://bsky.app/profile/justinhelps.bsky.social Twitter: https://twitter.com/primerlearning Music by Mathieu Keith. For business inquiries: mathieu.keith@proton.me Made with Godot Github: https://github.com/Primer-Learning/PrimerTools 0:00 Intro 1:11 A brief history and overview of artificial neurons 4:29 Training a neuron with one weight 8:00 A weight and a bias 14:28 Generalizing the update rule 17:55 Classifying mangoes by length 25:19 A thing that would not have worked 26:23 Classifying mangoes by length and mass 30:13 The exclusive OR problem 31:31 Credits and begging for money 32:26 Review