- What’s This? Just me pushing Scratch to its limits — testing if I can build a neural network (NN) in a visual coding sandbox. Spoiler: yes. Was it pretty? No. Was it awesome? Absolutely. - Instructions y_hat – The NN’s prediction. What it *thinks* the output should be. y – The actual/expected value. Reality check for `y_hat`. mse – Mean Squared Error. Measures how wrong the NN is, on average. Lower = better. Like golf, but with math pain.
Built by: @ScratchCat_000 (dev-kas) Powered by: XOR logic (because AND/OR were too basic) Digital chaos and sleepless ambition The forbidden art of neural nets on a visual code toy Acknowledgements: XOR is not a trademark. It’s a logical gate. It’s everywhere. Like caffeine. Scratch was not built for AI, but neither was my toaster. Here we are. Trademarks & Stuff: “Scratch” is a trademark of the Lifelong Kindergarten Group at the MIT Media Lab. Any names used (like XOR, AI, logic gates) are probably public domain or too old to trademark. This project is not affiliated with or endorsed by anyone except maybe my cat. All copyrights belong to their respective owners. I just copy the vibes. Fun fact: This XOR neural net runs on 100% pure visual block spaghetti.
