Toy Model v3.9.2 — Corridor Tearing Diagnostics and Metastable Phase Structure

Read full text
Authors: Hilmir Frímann Halldórsson, ChatGPT 5.2
Cited by: 0

Abstract

Toy Model v3.9.2 is a numerical laboratory for exploring metastability, phase coherence, and regime bifurcation in discrete recursive dynamical systems. This release introduces a consolidated diagnostic layer focused on corridor tearing phenomena—a class of behaviors in which trajectories exhibit long finite prefixes, structured outcome classes (e.g. escape vs grazing), and delayed instability governed by internal curvature rather than immediate divergence. Key features of v3.9.2 include: Seed-resolved and population-level diagnostics for exit time, regime classification, and phase evolution. Explicit separation between boundary location (parameter thresholds) and failure timing (time-to-instability). Curvature-based observables (e.g. second-difference diagnostics) designed to act as early-warning proxies. Fully reproducible CSV outputs supporting statistical analysis across large seed ensembles. This release is accompanied by the theory paper “Corridor Tearing Theory”, which formalizes the geometric and dynamical mechanisms diagnosed by the Toy Model and uses v3.9.2 as its primary numerical reference. The paper introduces a conceptual framework for interpreting metastable prefixes, phase commitment, and tearing events, while remaining model-agnostic beyond the diagnostics implemented here. Relationship to the theory paper The Toy Model provides the empirical substrate and reproducibility. The paper extracts and generalizes the theory implied by the diagnostics. Neither document supersedes the other; they are designed to be read together. This repository is intended for researchers interested in: metastability and delayed failure in discrete systems, geometry-driven phase transitions, diagnostic design for complex dynamical models, theory–simulation co-development workflows.

Fields of Study & Subjects

Probabilistic and Robust Engineering Design Nonlinear Dynamics and Pattern Formation Chaos Complexity and Education

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
2 January 2026