Skip to content

Void Dynamics Model (VDM): a physics‑first, axiomatized framework unifying discrete‑to‑continuum dynamics with zero‑shot learning & reasoning for real‑time agency. Lineage: AMN → FUM → VDM.

License

Notifications You must be signed in to change notification settings

justinlietz93/Prometheus_VDM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Void Dynamics Model (VDM)

A background‑independent, unified metriplectic field theory with an emergent causal cone and an epistemological J→M projection.

AMN-->FUM-->VDM

Current Status: Active development.

Author: Justin K. Lietz
Contact: [email protected]
Zenodo Community: Void Dynamics Model
ORCID: 0009-0008-9028-1366

Created: August 9, 2025
Last Updated: November 9, 2025

This research is protected under a dual-license to foster open academic
research while ensuring commercial applications are aligned with the project's ethical principles.
Commercial use requires written permission from the author.

Patent pending on specific AI technology - See license file for full terms

Static Badge

🔥News

  • November 10, 2025
    • Published results from the Counterfactual Echo Gain test to Zenodo
  • November 9, 2025
    • VDM will be upgrading to a dedicated custom physics engine that further enhances and improves scalability of the runner gating discipline, reporting, and automated unambiguous rule enforcement. This is being built into a fork of Void IDE as a code editor / computational physics lab.
  • November 6, 2025
    • 7 complete formalisms have been derived for VDM and documented as some of the final remaining pieces in the work to T8 Axiom 8 candidate.
  • November 4, 2025
    • Assisted‑Echo T4 (prereg v1c) rerun with instrument + plotting fixes:
      • Per‑seed plotting for A2/A5/B11 removes cross‑seed “seed‑chord” artifacts in overlays (error, ΔΣ, λ‑telemetry).
      • Noether (G1) tolerance scaled to instrument level: tol = max(1e-12, 10·eps·√N·max(h0,1)); routing remains governed by G1–G4 aggregation.
      • Reruns launched for N∈{512,1024}, dt∈{0.01,0.02,0.04}. Early results: N=512 dt=0.01 and N=512 walker dt=0.02 routed to main outputs with G1–G4 PASS; dt=0.04 routed to failed_runs pending CFL tightening.
      • Artifacts (PNG/CSV/JSON + figure packs) under Derivation/code/outputs/{figures,logs}/metriplectic, tag: assisted-echo-t4-prereg-v1c.
    • Next: upgrade G4 measurement to a multi‑point Δt fit (same thresholds) to harden RP‑1 decisions.
  • November 1, 2025
    • Published early draft of the proposed A8 Axiom candidate to Zenodo for provenance.
  • October 31, 2025
    • Added a T8 grade proposal for a new Axiom candidate as A8 - Lietz Infinity Conjecture which provides an elegant explanation for hierarchy and structure.
  • October 30, 2025
    • Added a historical/ folder including early original work like:
      • A self healing knowledge graph using Topological Data Analysis
      • As well as a Self Improvement Engine that integrates multiple reward components like novelty, self benefit, habituation, and TD error into a single "total reward" signal used to modulate its own neural plasticity for stable self-improvement that avoids weight saturation.
    • Validated the Counterfactual Echo Gain hypothesis by proving the trustworthiness and accuracy of the instrument, and showing that echo assist does modulate and improve the performance of a self aware system with 0 difference in cost compared to baseline.
  • October 28, 2025
  • October 23, 2025:
    • Created a sparsely populated CANON_PROGRESS.md document to post updates on private work to prevent this public repo from going stale.
  • September 29, 2025:
    • First public code release + creation of private Void Dynamics package which can now be imported and run in this repository using workflows and repository secrets.
  • September 28, 2025:
    • Posted two pre-prints to Zenodo. If you've published similar or relevant work on Reaction-Diffusion in the past 3 years on arXiv and want to support this work by endorsing me in a related category, submit an issue, post in the discussion board, or send me an email with Subject: RD Endorsement to get my attention. It would be much appreciated!
  • August 21, 2025: Launched public repo
  • ...
  • March, 2025
    • Released first falsifiable, reproducible simulations that validated initial claims.
  • February, 2025
    • SIE and Emergent KG + Real-Time TDA show very strong statistically significant findings, making feasible the entire model
  • October, 2024
    • Initial realization of the idea.

🎯Planned

  • Request endorsements to publish to a related category on ArXiv for latest validations.
  • Post recent validations as pre-print to Zenodo.
  • Publish a working demo of the VDM AI with a real-time self-organizing spiking neural network and self-healing emergent knowledge graph that performs topological data analysis on it's own graph during runtime.
  • Continue working toward other areas of Physics, like agency and determine how much the metriplectic findings affect other areas of physics
  • Once published to arXiv, take steps towards a peer reviewed journal and gather criticisms to refine my work.

DOIs

T4. Counterfactual Echo Gain (CEG): A Metriplectic Assisted-Echo Experiment Proposal in VDM DOI

T8. Axiom 8 Candidate: The Lietz Infinity Resolution Conjecture
DOI

A Logarithmic First Integral for the Logistic On Site Law in Void Dynamics
DOI

logo

❗NOTE

This organization is currently managed and operated by me (J. Lietz) alone as a solo developer / researcher. I may not respond by email right away. If you want to get my attention post in the discussion board briefly about what you'd like to talk about and let me know you sent me an email.

💬Discussion Board:


🧭What this is

This repository includes provides a public view of the Void Dynamics Model. It includes the theory, write-ups, code, notebooks, figures, logs, and validations for review by physicists, applied mathematicians, learners, students, and scientifically minded engineers.
Reproducible code released for the public is now available. Remaining proprietary work must be requested directly.

  • A set of derivation papers that establish a clean baseline physics slice starting from a discrete lattice, to a metriplectic split and cosmology.
  • Additional documents that explore future work (proposals).
  • Each paper separates what is proven from what is plausible or speculative and, where applicable, includes acceptance criteria for simple numerical checks.

🤖Why it relates to AI (brief)

  • The project studies how simple, but opposite metriplectic type rules yield stable, interpretable global behavior under resource constraints.
  • That design philosophy is relevant to AI systems that favor locality, event-driven updates, and transparent evaluation instead of opaque heuristics.
  • “Memory steering” (covered separately) frames slow routing bias and retention/decay as structured influences over faster dynamics, an analogy for directing computation without black-box shortcuts.

⚖️Licensing and scope

  • These materials are shared for academic review and discussion. Commercial use requires prior written permission. See the project’s license notice in the distribution or parent repository materials.
  • I reserve all legal rights to ownership of any custom or proprietary assets.
  • The scope stays within theoretical physics and simulation. Broad cosmological claims are withheld or clearly labeled until backed by derivation + numeric checks.

🔖Citations

📫Contact

  • For scope questions or clarifications about acceptance criteria, refer to the headers in the overview and topic files listed above. If you are reading this as part of a paper-only bundle, the maintainer’s contact is provided alongside the distribution materials.

About

Void Dynamics Model (VDM): a physics‑first, axiomatized framework unifying discrete‑to‑continuum dynamics with zero‑shot learning & reasoning for real‑time agency. Lineage: AMN → FUM → VDM.

Topics

Resources

License

Contributing

Stars

Watchers

Forks

Packages

No packages published

Contributors 3

  •  
  •  
  •