Skip to content

Zierax/Axiom-Zspace

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

38 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Before all else: I recognize that in a domain as complex and data-invariant as astrophysics, mathematical discrepancies or logical oversights may exist in these early results. Axiom-Zspace is a mission to build the first fully transparent, white-box system for exoplanet discovery. While it is a work in progress, I am committed to absolute transparency and invite the community to audit and correct any findings. I am open to technical collaborations under NDA to further refine these logical kernels. (Updated: 2/May/2026)

Axiom-Zspace v2.0: Deterministic Signal Deconvolution

License: MIT Python 3.8+ Status: Research-Phase

Project Overview

Axiom-Zspace is a high-performance, white-box engine designed for large-scale astronomical signal processing. By leveraging Deterministic Logical Invariants, the engine deconvolves raw light-curve data from the TESS and Kepler missions to identify exoplanet transit candidates with high precision and sub-millisecond inference logic.


April 26, 2026: First Milestone

We have successfully completed the processing of significant portions of TESS Sector 7 and Sector 41,42.

  • Result: 2886 New Exoplanet Candidates identified in 24-hour scanning. (Scans on going, 5000+ expected by end of April)
  • Verification: Each candidate is backed by a deterministic mathematical trace for individual validation.

May 2, 2026: Milestone Update

We have successfully completed the processing of significant portions of TESS Sector 7,36,41,42,55,67.

  • Result: 5845 New Exoplanet Candidates identified (Still on going).
  • Verification: Each candidate is backed by a deterministic mathematical trace for individual validation.

Performance Benchmark

The following benchmark represents a stress test of the Axiom-Zspace kernel on a constrained environment.

Metric Value
Hardware Environment Single-Core CPU
Total Processing Time ~4 Hours
Total Signals Tested 619
Confirmed Detections 574 (92.7%)
False Negatives (Missed) 3 (0.5%)
Technical Failures 42 (6.8%)

Note: Achieving a 92.7% detection rate on a single-core processor within 4 hours demonstrates the extreme computational efficiency of the Truthimatics logic compared to traditional probabilistic models.


Methodology & Open Invitation for Peer Review

This repository presents a white-box analysis. Unlike "black-box" neural networks, Axiom-Zspace utilizes traceable logic to isolate transit signals while minimizing noise interference.

Disclaimer: These findings are strictly classified as Candidates. While each discovery is supported by a deterministic benchmark and a unique analytical proof, they require spectroscopic confirmation by the professional astronomical community. To facilitate this, every entry includes its full analytical trace for independent verification.

I am a 17-year-old independent researcher currently in my final year of secondary school. As I am still in the learning phase of my journey, I invite astrophysicists, logic researchers, and data scientists to audit the underlying engine. My goal is to refine this framework through rigorous peer feedback, evolving it into a reliable, open-source tool for planetary discovery.


Special Acknowledgments

This project was made possible through the generous resources provided by the global research community.

Computational Support

A special thank you to the following platforms for providing the generous free-tier compute power that allowed me to "scan the skies" from a resource-constrained environment:

  • Google Colab
  • Kaggle Kernels
  • Lightning.ai

Data & Mission Support

  • NASA & The TESS Mission: For providing the high-fidelity, comprehensive public datasets that serve as the foundation for this research. Without this open-data initiative, independent research of this scale would be impossible.

Community & Validation Pathways

While Axiom-Zspace provides the deterministic proof, I highly value the human-in-the-loop validation approach. I am actively cross-referencing findings with community-driven projects.

  • Project Partner/Reference: Zooniverse Planet Hunters
  • Goal: Bridging the gap between high-frequency algorithmic detection and collaborative human verification.

Discussions


Contact & Collaboration: For technical audits or partnerships, please reach out via GitHub Issues or contact me directly at zs.01117875692@gmail.com.

About

Benchmarks on known planets 92.7% accuracy | Milestone(2/May/26): 5845 exoplanet candidate

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors