Fractions to Infinite P-adic Sequences
New interactive explorer showing how a rational fraction expands into an infinite sequence of p-adic digits.
P-adic Machine Learning
A single place for interactive demos, papers, and experiments on ultrametric machine learning. The current practical state-of-the-art here is: your systems, plus Zubarev's stochastic polynomial approach.
Interactive Demos
Theory + Results
Reproducible Projects
Explore Now
Working demos and projects you can click into immediately.
Hensel's Lemma Game
Hands-on lifting game for roots modulo powers of a prime.
P-adic Gibbs Regression Explorer
Interactive stochastic optimization and overfitting behavior for p-adic regression.
Regularisation Counterexample
Slide lambda and watch the fitted-point count increase when regularisation gets stronger.
P-adic Landscapes
Neighbour-hyperplane graph experiments and basin structure analysis.
State of the Art Snapshot
What works now
Practical explainable models and simulatable search procedures in ultrametric spaces, especially stepwise p-adic approaches and targeted stochastic walks.
What is hard
Optimal p-adic linear regression is NP-hard when dimension is part of the input, so local and structure-aware heuristics matter.
Key references
NP-hardness of p-adic linear regression
It's 2025 -- Narrative Learning is the new baseline to beat for explainable machine learning
Thesis-Derived Demos
Interactive notes and calculators derived directly from thesis chapters.
Parameter Instability Sandbox
Reproduce how tiny data changes flip coefficients while preserving external loss behavior.
NP-hardness Visualizer
Step through Max-Cut reduction into a 2-adic regression instance.
Hyperplane Contact Playground
Enumerate pair-defined lines and compare exact p-adic losses with rational arithmetic.
P-ademacher Mini Calculator
Explore residue-based complexity under random labels in base p.
Leading Zeros by Tag Rank
Interactive interpretation of valuation distributions in learned coefficients.
Loss Family Interpolation
Vary q in Lq = sum q^{-v_p(r_i)} and compare count-like, sum, and max-like regimes.
P-adic Series Convergence
See Euclidean divergence and p-adic convergence to -1 for geometric-style partial sums.
Metric Closeness Playground
Directly compare Euclidean and p-adic distances for rational pairs.
Zubarev Overfitting Lab
Interactive train/test curve simulator for degree and initialization choices.