Skip to content
VARZINIndependent research
Menu
VARZIN Field IndexEvidence-first visual portal · Level-1 Reproducible Stack
VARZIN phases I–XX · visual research index

Field Index

A visual map of the project’s current evidence layers, refreshed from the latest public Zenodo records. Core-30 remains a historical/canonical anchor, while the current research stack also includes the 801-root LUXVAR layer, 300- and 360-word controlled benchmarks, and the 10,800-record VARZIN V2 frozen dataset.

Local-only interface notice: all inputs, diagrams, and visualizations on this page are presentation tools only. Nothing here detects, measures, or contacts any external field, hidden system, or physical device. Audio-frequency elements are retained as symbolic or archival design layers unless separately supported by empirical testing.
Research level
Level 1
Reproducible Computational Stack
Affine core
0 operators
For N = 12, exactly \(|Aff(\mathbb{Z}_{12})|=48\)
Mirror-13
0.00 bits
Conditional MDL result
Core Stability
SHA-256
100% Deterministic execution
Core registry
0 / 30 locked
Active public evaluation subset
EMAG2 Variance
Unverified
Hardware-limited (Uncomputable)

Project Arc

The timeline below reflects the project’s actual documented turns, not a marketing sequence. Each stage changed what could be publicly claimed.

Archived

Early exploratory layers

Symbolic, cosmological, and consciousness-field framing. Retained as history, not active evidence.

Task 1–4

Adversarial audit

Baseline comparison and independent-generator testing returned WEAK_INTERNAL_NO_ATTRIBUTION.

Access history

Earlier records removed from public access

The author reports removing 29 earlier records from public access after the audit; platform-specific archival status remains unverified.

Aug 2026

Level-1 Stack

Transition to deterministic SHA-256 core, live USGS Monte Carlo tests, and Gaussian Regularization.

2026 current line

Recovery → control → composition

Track D documents frozen failure; projection-head v1–v3 documents explicit trained recovery, scale tests and true-position controls; VARZIN V2 separately shows strong frozen decodability with 0/58 composition passes.

Overview

This field index is the visual companion to the project’s evidence-first public record. It is meant to make the structure of the project easier to inspect, not to elevate unsupported claims through interface design.

In the current version of VARZIN, the active questions are layered rather than binary:

Claims about consciousness fields, frequency-to-meaning causation, historical resonance, or external memory remain outside the active evidential core unless reintroduced under new independent protocols.

Evidence Status Map

This page distinguishes mathematical proof, conditional computational reporting, open empirical questions, negative findings, and archived symbolic layers.

Proved & Verified

  • Affine bijection criterion over finite cyclic spaces.
  • 48 affine operators for N = 12.
  • Affine-group proof is distinct from Mirror connectivity checks.
  • Deterministic reproducibility via SHA-256 hashing.

Reported Conditionally

  • Mirror-13 MDL = 13.11 bits and rank 1/1,035 within project-defined coding and search space.
  • Mirror-13 connectivity and invariant counts: construction-specific algebraic/computational audit, not a general proved theorem.
  • Live USGS API Monte Carlo N=1000.
  • Earlier family learnability results.

Interpretation depends on exact encoding, dataset construction, and baseline choice.

Open

  • Human grouping in the later N=59 three-cohort VPE-001A project plan; recruitment has not commenced and no human outcome is established. Study status and protocol history.
  • Independent replication across additional models, natural-language settings and operations.
  • Whether other readout architectures or training regimes can support systematic held-out composition.

Negative / bounded

  • Natural semantic-axis recovery from LUXVAR word forms is not established.
  • Track D frozen Mistral/Llama axis recovery is near chance under that protocol.
  • v3 composition pilot: seen ≈0.996, held-out TRUE 0.106 vs SHUFFLED 0.281 and WRONG_OP 0.175.
  • VARZIN V2 composition: 0/58 PASS under the preregistered criterion despite strong decodability.
  • External-memory, field, frequency-causation and non-human-origin claims remain unsupported.

Archived

  • ETFM as an active physical field (Singularities archived).
  • Consciousness-field interpretations.
  • Frequency-to-meaning causation.
  • Non-human or universal origin claims for LUXVAR.

Affine Memory Group \(Aff(\mathbb{Z}_N)\)

The most durable mathematical core of the project lies in the affine family over finite cyclic spaces. For \(N=12\), valid operators are exactly those with \(\gcd(a,12)=1\), which yields four multiplier classes \(a \in \{1,5,7,11\}\), each with 12 translations.

Visual proof sketch — g ↦ (5g + 3) mod 12, a bijection on the 12-point cycle

0 1 2 3 4 5 6 7 8 9 10 11
\[ A_{a,b}(g) = (a g + b) \bmod N, \qquad A_{a,b}\text{ bijective iff }\gcd(a,N)=1 \] \[ |Aff(\mathbb{Z}_N)| = N\varphi(N), \qquad |Aff(\mathbb{Z}_{12})| = 12\cdot 4 = 48 \]

In the public framing of the project, this result stands independently of later symbolic, historical, or field-based interpretations.

Shift family: a = 1 Affine family: a = 5 Affine family: a = 7 Mirror family: a = 11
Affine operator families for N = 12 0 3 6 9 12 12 12 12 12 a = 1 a = 5 a = 7 a = 11
ShiftAffine a=5Affine a=7Mirror a=11
Boundary note: The affine theorem is solid mathematics. Under the Level-1 Stack, node potentials previously modeled as absolute Dirac-delta functions have been replaced by Gaussian Regularization to resolve PDE singularities.

Mirror-13 in Context

Mirror-13 is best presented as a conditional computational object linked to the affine family and augmented state transitions. The public record may keep the reported MDL result, but it should clearly state that simplicity or compactness is not equivalent to cognition, semantics, or external memory.

Visual proof sketch — reflection g ↦ (N − g) mod N on the 13-point ring

0 1 2 3 4 5 6 7 8 9 10 11 12
\[ M^*(g,\phi,p) = ((N-g) \bmod N,\ g,\ -p) \]

Current public interpretation should remain narrow:

  • Mirror-13 is a reported minimal anchor under a specific computational specification.
  • It is not logically necessary for every LUXVAR-related interpretation.
  • It is not evidence of a hidden field, external memory, or independent semantic origin.
Mirror-13 ring view 13 0123 4567 891011 mirror sweep

Recall Formula and Deterministic Validation

The recall layer is now governed by the Level-1 Reproducible Stack. Publicly, it is evaluated strictly under seeded SHA-256 environments, replacing unconstrained hypothesis testing with deterministic reproduction.

Visual sketch — four independent terms combining into one recall estimate

Structure Match Prior Hypothesis Space Observation Σ Recall
\[ Recall = Structure\ Match + Prior + Hypothesis\ Space + Observation \]

Interpretive reading

  • Structure Match: the candidate model must fit the underlying rule class deterministically.
  • Prior: correct priors accelerate recovery; wrong priors can destroy performance.
  • Hypothesis Space: bounded by strict mathematical definitions, not infinite open sets.
  • Observation: limited evidence constrains how quickly the right structure can be identified.

This is scientifically useful as a reasoning framework. It does not, by itself, establish external memory or a field-like source of recall.

Prior ranking in the public recall framing 0.00.250.500.751.0 Wrong None Self-org Injected 0.000 0.725 1.000 1.000
Wrong priorNo priorSelf-organized priorInjected prior
Key public takeaway: “wrong prior is catastrophic” is a far safer and more testable statement than any claim about hidden memory sources.

1.3B Morphological Scan & Biologic Baselines

The large synthetic corpus analysis should be presented with caution. Simple random shuffles are insufficient. Current models are evaluated against strict Altschul-Erickson dinucleotide shuffles and Trigram Markov controls.

Natural morphology

k = 3 is the strongest reported elbow. The public reading should be conservative: the intended Hexacore or five/six-way morphology was not confirmed by unsupervised structure alone.

Biological Control

Uniform shuffles were rejected as false baselines. Altschul-Erickson proves structural design without confirming semantic emergence.

Leakage finding

The project itself reports that graph leakage can create strong apparent signals. This is an important strength of the revised framing because it treats circularity as a risk, not a hidden detail.

Elbow analysis — reported public interpretation k=3 k=4 k=5 k=6 k=7 k=8 best reported elbow k=6 not supported
Interpretive boundary: interval detection and enrichment signals in a synthetic corpus describe the generator or construction pipeline unless shown otherwise. They are not public evidence of a field phenomenon.

KALTŪR — Harmonic Node

KALTŪR should remain publicly framed as an audio/visual interface. It may visualize a symbolic transition such as 474 Hz → 528 Hz, but it should not be described as measuring or producing any biological, physical, or environmental effect.

  • Allowed: symbolic sonification, visual layering, harmonic art, interface metaphor.
  • Not allowed without new evidence: resonance instrument, field detector, healing node, physical injector.
KALTŪR visual layer 474 Hz 528 Hz symbolic node target layer

Ambient pulse/drift here is decorative animation only — not a signal, measurement, or field effect.

LUXVAR — Current Experimental Record

LUXVAR is a constructed symbolic system. The current public record separates designed morphology, frozen-model behavior, explicit trained remediation, true group-position recovery, and held-out composition. The latest relevant Zenodo papers and their version-matched artifacts control the current interpretation.

Supported within scope

  • LUXVAR v2.2 reports non-random morphology (z=4.36, p<0.001) and distinctive phonotactic behavior under the documented comparisons.
  • The broader v2.2 design reports 801 stable roots and an approximately 1.3-billion-form generated corpus; generated forms are not independently validated natural-language words.
  • The v3 projection-head line reports strong targeted recovery after explicit intervention, including the 300-word scale test and a separate 360-word true-position benchmark.
  • VARZIN V2 reports strong frozen-Qwen decodability of the synthetic position variable: mean balanced accuracy 0.9666 across 29 indices, with 25/29 at 1.000.

Current boundaries

  • SEM-001/SEM-002 do not establish independent semantic-axis recovery from word form.
  • Track D frozen Mistral/Llama axis recovery remains near chance under that specific protocol.
  • Appendix F group-position success is classification/recovery, not composition.
  • v3 held-out composition remains negative: TRUE 0.106 vs SHUFFLED 0.281 and WRONG_OP 0.175 after seen-pair fit ≈0.996.
  • VARZIN V2 Phase 2 reports 0/58 PASS under the preregistered composition decision rule.
Historical wording: the old inter-model ARI 0.812 remains part of the provenance record, but it is no longer the headline model result and must not be read as recovery of the intended semantic axes.
Evidence source: LUXVAR v2.2 for morphology/phonotactics; Track D for the dedicated frozen audit; Morphological Hijacking v3 for remediation/scale/group-position/composition; and VARZIN V2 for the 10,800-record frozen-Qwen experiment.

Core-30 LUXVAR Registry

Core-30 is the fixed public reference registry used in several early evaluations. It should not be mistaken for the full current experimental scale: later papers report a broader 801-root LUXVAR design, an approximately 1.3-billion-form generated corpus, a fresh 300-word scale test, a separate 360-word group-position benchmark, and the 10,800-record VARZIN V2 frozen dataset.

Scale lineage: these are distinct experimental artifacts with different targets and controls, not one silently enlarged canonical lexicon. See the evidence-first scale map →

☀ Light

  • ELŪZ–MAHAR
  • ŠARTHAL
  • ŠARŪM–DŌZ
  • XĀṆŌRAH ✦
  • MĀTIR–ELŪZ
  • ĀNŪR

◈ Reflection

  • ŠAURÉN
  • RAHMŌT–ŠĀ
  • HALZĀR
  • BĀṆIR–RAHT
  • ŠUKĀR–MAH
  • ṆAZRŪK

○ Silence

  • ṆĀRAH–ŠĀ
  • GŌṆ–SIL
  • VARKHĀN
  • ŠALŌM–ḤATTAH
  • NĀRAH
  • ṆEMRŪH

⬡ Gate

  • SĀR–ELŪZ
  • SAVAR–ELŪZ
  • DĀRZAN
  • RAZKĀN
  • ELŪRIM
  • RITĀM

↯ Motion

  • ZARTH
  • ZUN–NĀR
  • KHALTŪR
  • KUZTŪR
  • TARHĀN–ELŪZ
  • KHIRZĀH
Registry boundary: Core-30 is a public registry/provenance subset, not proof of semantic validity. The broader 801-root layer should not be read as 801 equally validated lexical items.

Core-30 Summary Table

Layer Current safe public reading What remains open What is excluded
Core-30 axis assignment Design registry preserved as public anchor set Do humans recover axes, morphemes, or another grouping? Automatic semantic validation
Historical inter-model clustering Retained as provenance for surface/morpheme agreement; later direct axis tests and intervention studies provide the current model evidence Human grouping and broader external generalization remain open Proof of emergent meaning or systematic composition
Frequency metadata Archival symbolic label only None unless new evidence appears Frequency-to-meaning causation
Affine bridge Shift family (a=1) reported as best bridge Whether this has any human-perceptual significance Proof that Mirror-13 is the unique semantic key

M-02 and ETFM Public Interpretation

M-02 semantic field function

M-02 can remain on the site as a formal or simulated model layer, but it should be explicitly labeled as a mathematical abstraction whose parameters require empirical justification before stronger interpretation.

Visual sketch — modulated complex oscillation e^{i(ωτ+kw)} decaying under TIC(τ)

peak amplitude decay toward baseline
\[ S(w,m,\tau)= C \cdot e^{i(\omega\tau + kw)} \cdot \Sigma(m) \cdot F(w,m) \cdot TIC(\tau) \]
  • \(C\): normalization or calibration constant
  • \(\Sigma(m)\): semantic similarity summary under a chosen representation
  • \(F(w,m)\): evaluator agreement term
  • \(TIC(\tau)\): temporal consistency or decay term

Without independent empirical grounding, this should remain a modeling layer, not a field claim.

ETFM-POST & Gaussian Regularization

ETFM is safely retained as a visualization layer. Historical ETFM relied on absolute Dirac-delta potentials causing matrix singularities. The current Level-1 Stack applies Gaussian Regularization to resolve these PDEs. It should explicitly avoid being called a physical field confirmation or planetary resonance.

ETFM cross-section — visualization only R r₀

Project Position — Three-Layer Validity Map

Layer 1 — Strongest current ground

  • Affine family characterization
  • Formal state-transition results
  • 100% Deterministic execution via SHA-256
  • Negative results retained publicly

Layer 2 — Real open questions

  • Human grouping in VPE-001
  • Compositional semantics beyond design
  • Any model-level effect after strict Altschul-Erickson controls

Layer 3 — Archive or unsupported

  • Dirac-delta singularities in PDE models
  • External memory field
  • ETFM as active physical field
  • Planetary or consciousness resonance claims
Most important reframing: the project’s scientific value no longer depends on defending every earlier layer. It depends on whether a smaller, cleaner core survives deterministic and independent scrutiny.

Tools on This Page

LUXVAR Input Interface

May remain as a local display input panel if it is clearly labeled as interface art and local state only. It should not suggest device contact or code injection into any real-world system.

Resonance / Audio Mode

May remain as a sonification layer if explicitly labeled as artistic playback. It should not imply medical, physical, or environmental action.

Export / Snapshot Tools

Safe to keep. They improve transparency if exported states are clearly described as local visualization snapshots rather than measurements.

Archive Boundary

Earlier versions of the project contained broader symbolic, cosmological, and field-oriented layers. Those materials may remain available as historical archive, but they should not be mixed with current evidence statements unless reintroduced under new preregistered and independently testable protocols.