Physics Simulation

Text lets an incorrect explanation remain elegant. A simulation is less polite. The bridge falls, the orbit escapes, or the ball passes through the floor. I wanted to test a narrow version of a larger idea: Can a model with fewer than one billion parameters learn enough structured code to generate small interactive physics worlds? The repository contains a completed Qwen3 0.6B LoRA run built from synthetic p5.js examples. It also contains a developmental learning proposal for MuJoCo....

What happens when you take one of the smallest language models available, feed it a thousand physics animations generated by one of the largest, and ask it to teach K-12 students about science? You get a model that weighs less than a gigabyte, trains in under 3 minutes, and generates interactive physics simulations on demand. The Premise LLMs are getting bigger. GPT-5, Claude Opus, Gemini Ultra – they’re all racing to hundreds of billions of parameters....

What if AI agents learned about the world the way babies do—by touching, tasting, dropping, and breaking things? When a toddler drops a spoon for the 47th time, they’re not being annoying. They’re conducting physics experiments: testing gravity, observing bounce patterns, mapping cause and effect. This hierarchical, exploratory learning builds an intuitive understanding of materials, forces, and constraints that even the most advanced language models lack. The gap is becoming increasingly obvious: LLMs can write eloquently about physics but don’t truly understand that dropping a glass causes it to shatter, or that wet surfaces are slippery....