v1.0.0 Stable Release

Contractive Recurrence.
Certified Convergence.

PIRTM is a Python library for prime-indexed recursive tensor mathematics with provable stability bounds, adaptive governance, and ethical gating.

Contraction-guaranteed
CSL-compliant
Audit-chained
Recurrence Equation
Xt+1 = P(ΞtXt + ΛtT(Xt) + Gt)
Ξt : State Projection
Λt : Adaptive Operator
T(X) : Tensor Transform
Gt : Governance Gate

Precision for High-Stakes Recursion

PIRTM is not a general-purpose numerical solver. It is a specialized engine for prime-indexed recursive systems where stability is non-negotiable. Unlike standard solvers that drift, PIRTM enforces contractive dynamics at every step.

Designed for applied mathematicians and numerical engineers, it provides a rigorous framework for running recursive tensor operations with mathematically guaranteed convergence bounds.

Whether you are modeling quantum cognition or complex adaptive systems, PIRTM ensures your recurrence remains bounded, auditable, and ethically gated through the CSL framework.

Contractive Step

Every step satisfies q < 1 - ε, ensuring strict convergence towards a stable fixed point.

Adaptive Margin

Epsilon auto-tunes via residual + contraction telemetry to maximize stability.

ACE Certification

Post-run certificates with tail bounds and ISS guarantees for full auditability.

Spectral Analysis

Prime-indexed eigenvalue decomposition and phase coherence tracking.

The PIRTM Pipeline

01

Define Operators

Supply T, P, and parameter sequences Ξt, Λt, Gt to configure the system dynamics.

02

Run Recurrence

pirtm.run() auto-projects parameters to maintain contraction and records telemetry per step.

03

Certify Results

ace_certificate() produces a Certificate with margin, tail bound, and convergence status.

Developer Interface

Every step is telemetered. Every run is certifiable.
import numpy as np
from pirtm import step, ace_certificate

# Initialize State and Parameters
X = np.zeros(4)
Xi = 0.3 * np.eye(4)
Lam = 0.2 * np.eye(4)
T = lambda x: np.tanh(x)
G = np.random.randn(4) * 0.01
P = lambda x: np.clip(x, -1, 1)

# Run a single step with epsilon constraint
X_next, info = step(X, Xi, Lam, T, G, P, epsilon=0.05)

# Generate Audit Certificate
cert = ace_certificate(info, tail_norm=0.01)

print(f"q={info.q:.4f}, certified={cert.certified}, margin={cert.margin:.4f}")

Tier Roadmap

Tier 1: Installable PackageComplete

Core library structure and pip distribution.

Tier 2: Complete API SurfaceComplete

Full implementation of step, run, and certify functions.

Tier 6: Emission GatingIn Progress

Telemetry + Q-ARI integration for safety.

Tier 8: Ξ-CertificationPlanned

Full pipeline for constitutional certification.

The Multiplicity Ecosystem

PIRTMRecurrenceΛ-ProofGovernanceDRMMFeedbackΞ-Constitution

Start building with PIRTM

pip install pirtm
View Documentation