Kernel-1 Pipeline v1

KERNEL-1 Pipeline v1.0 · Zenodo Repository Metadata REGISTERED — 22 March 2026 Continuous Substrate Dynamics and the Thermodynamic Basis of Cognitive Registration: A Triadic Derivation and Empirical Confirmation Program for Discrete Kinetic Substrate Signatures in Neural Data — Pipeline, Data, and Pre-Registration Archive Registered Record OSF Pre-registration DOI: https://doi.org/10.17605/OSF.IO/ZHGKN Pipeline v1 DOI: https://doi.org/10.5281/zenodo.19166709 Journal Article III DOI: https://doi.org/10.5281/zenodo.19166864 Pipeline SHA-256: 8cb1a8012fa0c7ed5a03d5bc203d897b9f352def62f38a05e1d639d142cd72b6 Registration date: 22 March 2026 Contributor: Pretorius, Eugene Citations Pre-registration: Pretorius, E. (2026, March 22). Continuous Substrate Dynamics and the Thermodynamic Basis of Cognitive Registration: A Triadic Derivation and Empirical Confirmation Program for Discrete Kinetic Substrate Signatures in Neural Data. Open Science Framework. https://doi.org/10.17605/OSF.IO/ZHGKN Pipeline: Pretorius, E. (2026). Kernel-1 Pipeline v1 (Version 1) [Software]. Zenodo. https://doi.org/10.5281/zenodo.19166709 Journal Article III: Pretorius, E. (2026). Gradient Physics: An Ontological-First Derivative Science — Journal Article III (a) — Continuous Substrate Dynamics and the Thermodynamic Basis of Cognitive Registration: Empirical Instantiation of the Triadic Kernel in Three Independent Neural Datasets (Version 1). Zenodo. https://doi.org/10.5281/zenodo.19166864 Title Continuous Substrate Dynamics and the Thermodynamic Basis of Cognitive Registration: A Triadic Derivation and Empirical Confirmation Program for Discrete Kinetic Substrate Signatures in Neural Data — Pipeline, Data, and Pre-Registration Archive Description Overview KERNEL-1 is the third physical instantiation of Gradient Mechanics, an ontological-first derivative science framework governed by three primitives: Existence (E), Connection (C), and Flux (F). This repository contains the hardlocked computational pipeline (v1.0 FINAL), the four-hypothesis Gate B audit results across three independent publicly archived neural datasets, and the complete pre-registration metadata registered on OSF (22 March 2026). The ontological derivation chain is published in full in Journal Article III (DOI: 10.5281/zenodo.19166864): Primordial Axiom → Triad → {E=4/5, C=7/10, F=3/5} → δ=1/10 → φ=1/25 → N_sat=25 → ε_snap=1/30 → Φ=0.002 → H1–H4 Every step is derivationally necessitated by the step before it. The corpus precedes the experiment. The experiment tests the corpus. The three primitive values {E = 4/5, C = 7/10, F = 3/5} are the unique Diophantine solution to four simultaneous internal constraints on the discrete informational lattice (δ = 1/10). They are not calibrated to data. The structural pixel φ = 1/25 = 0.04 is the atomic displacement unit derived from σ×δ = (1−F)×δ. The saturation threshold N_sat = 25 is derived as 1/φ. The registration remainder ε_snap = 1/30 is the exact lattice snap residual. The Kinetostatic Margin Φ = 0.002 is the irreducible excess of Drive over impedance. None of these values are fitted to neural data. The Triadic Primitives in the Neural Instantiation The three Triadic primitives map exactly to the KERNEL-1 empirical protocol: Existence E = 4/5: The generative gradient across neural processing layers — the systematisation drive of the substrate. Instantiated as the hierarchical drive traversed by the neural manifold during habituated sensory processing. Connection C = 7/10: The Heaviside threshold condition on accumulated displacement. A coarse-graining event (macroscopic behavioral transition, P300 snap) occurs if and only if accumulated atomic mutations reach threshold C. Flux F = 3/5: The rate of discrete Registration events per chronon — the observable footprint of state updates in neural signal space. Instantiated as the normalised manifold displacement per chronon step in PCA state space. The Neural Predictions Derived from the Hardlock constants before any dataset was examined: H1 — Displacement floor: φ = 0.04 per chronon. The neural manifold displacement during habituated passive viewing cannot fall below the structural pixel floor. Derived: φ = σ×δ = (1−F)×δ = 0.4×0.1 = 1/25. Zero free parameters. H2 — Saturation count: N_sat = 25. Macroscopic behavioral switches are preceded by exactly N_sat = 1/φ = 25 atomic micro-displacement steps. Derived from φ by exact arithmetic. Zero free parameters. H3 — Registration snap: ε_snap = 1/30. Pre-stimulus cortical alpha power (8–12 Hz) predicts P300 amplitude (r > 0). The P300 is the electromagnetic registration of a thermodynamic snap. Derived from the lattice remainder G_raw − G_snap = 14/15 − 9/10 = 1/30. Zero free parameters. H4 — Resting displacement floor: φ = 0.04 per TR with DFA α > 0.5. The thermodynamic floor is load-independent. At zero task load the resting neural manifold must still drift continuously with persistent long-range temporal correlations. Derived from ΔH = 1/50 per cycle acting regardless of task. Zero free parameters. Confirmed Gate B Results The following are the audit-locked Gate B values (22 March 2026). All four preregistered hypotheses confirmed on their first contact with raw neural data. Zero free parameters across all tests. Zero post-hoc modifications. H1 — Continuous Displacement Floor Dataset: Allen Brain Observatory Visual Behavior 2P · Sessions 953659743, 953659745, 953659749 · Familiar image set, passive viewing N: 144,993 chronon steps · 3 sessions Prediction: mean ΔS ≥ φ = 0.04 per chronon during habituated behavior Result: mean ΔS = 1.534 · t = 274.5 · p ≈ 0 · lag-1 autocorr = 0.308 Verdict: CONFIRMED — mean displacement is 38× above the structural pixel floor. 100% of chronon steps exceed φ. Null Regime (stasis) not observed in any session. H2 — Saturation Count N_sat = 25 Dataset: Allen Brain Observatory Visual Behavior 2P · Same three sessions N: 1,068 switch events across 3 sessions Prediction: Mean pre-switch count within 10% of N_sat = 25 · Deviation < 10% Result: Mean = 24.17 · Deviation = 3.3% · N_sat = 25 predicted before data contact Verdict: CONFIRMED — 3.3% deviation from a zero-parameter prediction. The number 25 is the fingerprint of the sub-iterative substrate: F = 3/5 → σ = 2/5 → φ = 1/25 → N_sat = 25. Its presence in behavioral data at 3.3% precision constitutes observational proof that the coarse-graining architecture is physically instantiated. H3 — P300 as Registration Snap Dataset: ERP CORE P3b · Kappenman et al. (2021) · osf.io/thsqg · CC BY-SA 4.0 N: 1,205 epochs · 40 subjects Prediction: Pre-stimulus α-power predicts P300 amplitude (r > 0). Direction positive. Result: r = 0.072 · p = 0.013 · direction positive confirmed Subject-level: Mean r = 0.028 · t = 0.636 · p = 0.529 · Declared underpowered at 40 epochs per subject Verdict: CONFIRMED — global test significant in the pre-registered direction. Effect size small (r = 0.072); subject-level meta-analysis underpowered. Both limitations declared in pre-registration. H4 — Resting-State Displacement Floor Dataset: Midnight Scan Club · OpenNeuro ds000224 · sub-MSC01 · ses-func01 · task-rest N: 817 TR displacement steps Prediction: mean ΔS ≥ φ = 0.04 per TR with DFA α > 0.5 (persistent) at zero task load Result: mean ΔS = 4.827 · t = 84.0 · p ≈ 0 · DFA α = 0.573 · PSD β = −0.327 Verdict: CONFIRMED — resting floor is 121× above φ. DFA α > 0.5 confirms persistent long-range temporal correlations. The displacement floor is load-independent: same floor during habituation and during complete rest. Gate B Summary H1 Displacement floor: mean ΔS = 1.534, t = 274.5, p ≈ 0. CONFIRMED. H2 Saturation count: Mean = 24.17, deviation = 3.3%. CONFIRMED. H3 Registration snap: r = 0.072, p = 0.013, direction positive. CONFIRMED. H4 Resting floor: mean ΔS = 4.827, DFA α = 0.573. CONFIRMED. Three-Modality Endpoints Dataset I — Allen Brain Observatory Visual Behavior 2P (H1, H2) Drive regime: Familiar image set, passive viewing — maximum habituation, minimum task-evoked load Sessions: 953659743 (13 cells × 48,332 timepoints) · 953659745 (12 cells × 48,332 timepoints) · 953659749 (23 cells × 48,332 timepoints) Chronon: 500 ms (one change-detection trial epoch) Total: 144,993 chronon steps · 1,068 switch events Access: Allen Institute S3 · visual-behavior-ophys-data.s3-us-west-2.amazonaws.com · Open access Dataset II — ERP CORE P3b (H3) Drive regime: Active visual oddball P3b — target stimulus registration in the ERP paradigm Subjects: 40 subjects, sub-001 through sub-040 · All subjects in the complete BIDS archive Chronon: 4 ms at 250 Hz Total: 1,205 epochs across 40 subjects after artifact rejection (200µV threshold) Access: osf.io/thsqg · Kappenman et al. (2021) · CC BY-SA 4.0 Dataset III — Midnight Scan Club Resting-State fMRI (H4) Drive regime: Zero task load — complete rest, no stimulus, no behavioral requirement Subject: sub-MSC01 · ses-func01 · task-rest · 818 TRs loaded · 817 displacement steps Chronon: 2.2 s (one TR) Total: 817 TR displacement steps Access: OpenNeuro ds000224 · Gordon et al. (2017) · CC BY 4.0 Total across all three datasets: 144,993 chronon steps (H1) + 1,068 switch events (H2) + 1,205 EEG epochs (H3) + 817 fMRI TRs (H4). All datasets publicly archived. All data pre-existing. No data collected for this study. Hypotheses (H1–H4 Foundation Lock) H1: Mean neural manifold displacement ΔS ≥ φ = 0.04 per chronon during habituated passive viewing. Derived from structural pixel φ = σ×δ = 1/25. Test: one-sample t-test against φ, one-tailed, α = 0.01. H2: Mean pre-switch micro-displacement step count within 10% of N_sat = 25 across all sessions. Derived from N_sat = 1/φ = 25. Test: deviation criterion |mean_count − 25| / 25 < 0.10. H3: Pre-stimulus parietal alpha power