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 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.
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.
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.
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.
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.
This project was made possible through the generous resources provided by the global research community.
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:
- 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.
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.
- https://www.zooniverse.org/projects/nora-dot-eisner/planet-hunters-tess/talk/2110/3998656
- https://www.zooniverse.org/projects/nora-dot-eisner/planet-hunters-tess/talk/2110/3995685
- https://www.zooniverse.org/projects/nora-dot-eisner/planet-hunters-tess/talk/2110/3987225
Contact & Collaboration:
For technical audits or partnerships, please reach out via GitHub Issues or contact me directly at zs.01117875692@gmail.com.