Computational Notebooks

Two reproduction notebooks regenerate every numerical result in the framework live from the ppm Python package. A third demonstration notebook runs the experience-binding mechanism end-to-end as a tunable simulation. View them rendered on GitHub, or run them interactively in the browser via Voila on Fly.io — no install required.

View Repository on GitHub ▶ Launch on fly.dev Ask in NotebookLM Open the Model
01

Predictions and Sensitivity

Every numerical prediction PPM makes, computed live from the ppm Python package — no hardcoded values. Inline sliders at the load-bearing inputs let you move structural choices and watch the predictions break.

View on GitHub nbviewer Run ▶
02

Verification Ledger

The framework's signature numbers — α, g, G, Λ, the Jarlskog J, the mass hierarchy — derived from CP³ and the pion mass step by step. Each cell does the math from primitives; a self-consistency suite at the end checks the whole package.

View on GitHub nbviewer Run ▶
03

Experience Binding Demonstration

N sub-stream agents searching the same V₄ measurement torus for the viewing frame that resolves a hidden signal buried in noisy channels. One group is bound through shared-environment coupling; the other runs the same task in isolation. Sliders let you adjust witness count, coupling strength, per-witness noise, and the late-stage learning floor, then re-run the simulation. Bound agents converge together and track the signal; individual agents stay scattered and don't.

View on GitHub nbviewer Run ▶