Why this one matters. Every machine-learning model, from this straight line to a large language model, repeats the same three steps: make a prediction, measure how wrong it is, then nudge its numbers to be less wrong. This page is that loop at its smallest — only two numbers to learn, a0 and a1, so you can actually watch it happen.
Stuck on a label? Any label followed by a small mark explains itself — try this one and the explanation opens underneath it. Click again to close. For the formulas, open 📝 Notes at the top of the page.
Arrived from the home page? There you placed a line by hand. Here the machine finds one by itself. Press Initialize, then Start, and watch the red error bars shrink.
Everything here is yours to choose. Everything below is found by the machine, one small correction at a time.
The line stopped moving because the error stopped shrinking. Loss fell from — to —. Those red bars are exactly what it was minimising — one per data point, each the gap between the real value and the prediction.