NN-1 · Core lab · 35–45 min
Make it fail, then make it learn
Use the Tiny Neural Network to investigate linearity, hidden-layer capacity, training dynamics, and generalization by deliberately creating both failure and success cases.
Open NN-1AI Playgrounds · Activity Packs
These pilot activities turn an interactive applet into a complete learner task: predict → run → observe → explain → transfer. Student responses stay in the browser and can be printed. No account or backend is required.
NN-1 · Core lab · 35–45 min
Use the Tiny Neural Network to investigate linearity, hidden-layer capacity, training dynamics, and generalization by deliberately creating both failure and success cases.
Open NN-1CNN-1 · Core lab · 40–50 min
Use the Convolution Playground to connect hand-calculated multiply-and-sum arithmetic to edge detection, learned filters, and pooling.
Open CNN-1These public pages are student-facing. Teacher answer keys and grading exemplars are intentionally not published on the public site. The current Activity Pack pilot is English-only; the core applets remain multilingual.