START HERE · ENGLISH
A clear route into the research
Choose a practical starting point, understand the evidence classes, and find the exact sources behind a result.
1. Understand the project
A signal is a sequence of measurements, such as vibration over time. A graph describes elements and their connections. A causal order describes which events precede others. X6 explores methods for analysing these mathematical objects and checking their results.
A spectrum is a collection of values derived from a matrix. Different structures can share a spectrum; the project includes explicit counterexamples. A pattern in an image alone is not evidence for prime numbers or a new physical mechanism.
The current priority is Lorentzian spectral geometry. Signal processing, RAX, photonics and industrial data studies have their own scopes and acceptance gates. There is no claim of a completed quantum computer or a proven quantum-gravity theory.
2. Try a browser experiment
Open Signal Validation Lab. Start with a synthetic example whose clean signal is known. Compare the original signal, a noisy version and the reference-filter outputs. Inspect reconstruction errors as well as spectral summaries.
Explore the documented counterexamples, including polarity inversion: a signal and its negative can share a power spectrum. Similar-looking spectral summaries do not establish faithful signal recovery.
Export the experiment evidence and use the documented import/replay workflow to check the calculation. A successful replay shows computational consistency within that workflow; it does not prove data authenticity or physical device performance.
3. Reproduce a reference calculation
For the common offline reference suite, install Git, Python 3.12 and Node.js 24 first. Cloning requires network access; the reference computation itself needs no package download or new Wolfram call. Use a fresh output directory.
The commands below reproduce the current main checkout. For a versioned result, use the reference-edition page and its pinned source commit or immutable release package. Record the commit and runtime versions in any report.
Expected status: PASS_OFFLINE_REFERENCE_REPLAY. Read the individual report scopes and rejection controls. Reproducing a negative or inconclusive scientific result can be the correct software outcome.
git clone --branch main https://github.com/thomasmgfeller/x6-spectral-astrolabium.git x6
cd x6
python3 -S -B x6.py reference qualify --output /tmp/x6-reference-evidence4. Interpret status and evidence
MAIN means the registered source is available in the main branch. BRANCH_ONLY means the documented research source requires the named branch and pinned commit. These are source-location labels, not quality ratings.
SOFTWARE means a bounded executable checking path is documented. CONCEPT and DOSSIER describe plans or documents without a standalone executable whole. A photonics module can be SOFTWARE because its digital reference is testable while the physical device remains unqualified.
An exact finite computation concerns its specified finite objects. A numerical simulation concerns its model and assumptions. Archived laboratory data are measurements from their original acquisition context. None of these automatically establish a new physical measurement, industrial superiority or independent external review.
PASS, FAIL, NOT_RUN, NOT_READY and INCONCLUSIVE have different meanings. Read the stated claim, the evidence report and the next acceptance gate together. Each evidence report retains its own checked revision; a newer source pin does not retroactively qualify changed code.
5. Cite, reuse and contribute
Cite the precise release or commit you used, and cite external datasets separately. CITATION.cff describes the selected finite reference edition; it is not a certification of every module.
Original project contributions use Apache-2.0. Separately licensed components and datasets keep their own conditions. Included Paderborn data retain CC BY-NC 4.0; the portal does not grant commercial rights to those data.
Useful contributions include independent reproductions, counterexamples, accessibility improvements and clearer documentation. The optional llms.txt and JSON index help navigation; they do not guarantee indexing, ranking or inclusion in an AI answer.