Andrei Ursachi · Bibliothèque de recherche
La recherche, en détail.
Explorez les questions, les méthodes et les travaux en cours derrière les articles. Recherche publiée, prépublications et brouillons présentent leurs sources originales et leur statut de publication.
Recherche publiée
Article évalué par les pairs · Frontiers in Human Neuroscience ·
Golden ratio organization in human EEG is associated with theta-alpha frequency convergence: a multi-dataset validation study
Studies the association between theta–alpha frequency organization and convergence near the 8 Hz boundary in resting-state human EEG.
Démarche
Spectral-centroid analysis of the public PhysioNet EEGBCI and LEMON datasets, with null-model comparisons, replication and sensitivity checks.
Portée et questions ouvertes
An observational association in the analysed datasets. It does not establish a causal mechanism or a universal organizing law.
Programmes de recherche
Prépublication · Zenodo ·
The Riemann Hypothesis and Yang–Mills Theory: Shared Operator Methods and Open Bridges
Une comparaison d’estimations d’opérateurs pour l’hypothèse de Riemann et le modèle SU(2) sur réseau. La prépublication présente un certificat pour la forme arithmétique régularisée et une estimation sur réseau, puis précise les conditions encore ouvertes.
Démarche
The manuscript keeps equality modes, signed responses and complete tails in their respective operator formulations. The latest revision treats a regularized arithmetic form at a specified cutoff and a family of SU(2) lattice parameters.
Portée et questions ouvertes
A zero-tolerance sign bound along an unbounded sequence of cutoffs remains open. The lattice analysis leaves a conditional comparison to prove and a separate passage to the continuum with uniform decay in physical time. The paper is a preprint.
Prépublication · Zenodo ·
Local Weil Positivity and Signed Prime-Deficit Criteria
Un certificat local de positivité et des critères arithmétiques signés pour étudier l’hypothèse de Riemann, avec la condition restante à toutes les échelles explicitement formulée.
Démarche
Computer-assisted control of a complete infinite-dimensional complement, signed prime-power deficits and an explicit formula retaining all zeros.
Portée et questions ouvertes
The remaining all-scale condition is explicit and open. The local certificate and finite computations do not constitute a proof of the Riemann hypothesis.
Document de travail · Zenodo ·
Spectral-gap certificates and signed response for the two-dimensional O(3) model
Certificats de gap spectral à paramètres finis et réponse signée dans le modèle O(3) bidimensionnel, avec des hypothèses explicites et des questions ouvertes pour la suite.
Démarche
Finite-purity estimates, relative-rotor and graph-determinant analysis, and spherical elimination under stated comparison and arithmetic premises.
Portée et questions ouvertes
Results concern this two-dimensional model and the stated regimes. They do not establish a four-dimensional Yang–Mills mass gap.
Prépublications et manuscrits
Les prépublications et manuscrits sont partagés pour examen et n’ont pas été évalués par les pairs.
Prépublication · Zenodo ·
Azimuthal finite-difference identities for the slow-spin leading-eikonal Kerr damping rate
Derives azimuthal finite-difference identities for slow-spin Kerr quasinormal-mode damping within the leading-eikonal expansion.
Démarche
An ordered fixed-m re-expansion is compared with a finite Richardson grid of gravitational modes.
Portée et questions ouvertes
The identities retain their specified spin and eikonal order; numerical agreement is a finite-regime diagnostic.
Prépublication · Zenodo ·
Sound timing and visual reports in a public behavioral dataset
Reanalyses whether sound timing changes visual reports in an existing public behavioural dataset.
Démarche
Participant-level signed-error contrasts, cleaning checks, bootstrap intervals and sign-flip comparisons.
Portée et questions ouvertes
A secondary-data reanalysis of an interaction already reported by the source investigators, with its own cleaning and analysis choices.
Prépublication · Zenodo ·
Ordinary spectra can mimic a Schumann anchored golden ratio enrichment pattern
Tests whether a preferred frequency anchor uniquely identifies golden-ratio organization in EEG-like spectra.
Démarche
Analytic calculations and synthetic spectra compare ordinary rhythm peaks, aperiodic backgrounds and detector behaviour.
Portée et questions ouvertes
Ordinary spectra can reproduce parts of the detector signature. Anchor selection alone does not identify a generating mechanism.
Prépublication · Zenodo ·
Limited incremental prediction of cardiac synchrony judgments from heartbeat locked scalp activity
Asks whether heartbeat-locked scalp features improve prediction of later cardiac synchrony judgments beyond other covariates.
Démarche
Regularized logistic comparisons on 1,731 omission trials from an existing 34-person archive, with participant-held-out labels.
Portée et questions ouvertes
Outcome-blind calibration used full participant waveforms, so the reported evaluation is offline rather than a deployment simulation.
Prépublication · Zenodo ·
Signed zeta potential in the LNP Atlas website export
Audits the consistency of signed zeta-potential text and a numeric helper in a pinned LNP Atlas export.
Démarche
Qualifier-preserving parsing, row-level sign comparisons and comparison with a separately downloaded versioned artifact.
Portée et questions ouvertes
The issue concerns the inspected export and helper field. It is not a general conclusion about the underlying experiments.
Prépublication · Zenodo ·
Auditory distraction and report dynamics in two behavioral archives
Examines whether group auditory-distraction effects also support stable individual differences across tasks.
Démarche
Separate reanalyses of two behavioural archives, with cross-task rank comparisons and an explicitly retrospective analysis.
Portée et questions ouvertes
The primary cross-task rank test did not pass. Group effects and individual-difference claims remain distinct.
Prépublication · Zenodo ·
Identifying a persistent shared state in a synthetic intervention benchmark
Defines an operational signature for a persistent shared state using delayed influence, erasure and address specificity.
Démarche
Finite synthetic benchmarks compare a planted persistent state against five specified alternatives and test retention recovery.
Portée et questions ouvertes
One initial retention-recovery gate failed. A synthetic discrimination result does not establish a physical or biological shared state.
Prépublication · Zenodo ·
Exact finite comparisons of a symmetry orbit and uniform binary constraint relations
Tests how the sampling law of a random constraint relation affects joint inclusion and second moments.
Démarche
Exact arithmetic on a finite four-value symmetry example compares a permutation orbit with uniform relation sampling.
Portée et questions ouvertes
The conclusions are exact for the stated finite constructions; equal density and overlap do not make the probability laws identical.
Prépublication · Zenodo ·
Finite-multipole damping order and endpoint selection in Schwarzschild quasinormal-mode pairs
Studies how finite-multipole Schwarzschild damping approaches its leading-eikonal limit and selects pairwise endpoints.
Démarche
A sector- and overtone-dependent sign threshold is checked against a finite collection of computed modes.
Portée et questions ouvertes
The endpoint conclusion applies within windows satisfying the specified order; finite confirmation is not a result for every multipole.
Prépublication · Zenodo ·
Adaptive feature acquisition under diagnostic misspecification
Examines when adaptive feature acquisition helps under complementarity, redundancy and diagnostic misspecification.
Démarche
A fully specified finite Bayesian model compares a greedy conditional-information policy with an optimized static pair.
Portée et questions ouvertes
Results depend on the constructed worlds, observation budget and diagnostic model; they are not a universal advantage for adaptivity.
Prépublication · Zenodo ·
Accuracy retention and width allocation at fixed parameter count: A reproducibility correction to The φ-Compression Law
Corrects the interpretation of an efficiency-retention identity previously attributed to golden-ratio network widths.
Démarche
An executable source audit and fixed-parameter CNN comparisons separate algebraic normalization from trained accuracy.
Portée et questions ouvertes
The identity reduces to normalized accuracy for every width schedule. It does not establish a sequence-specific conservation law.
Prépublication · Zenodo ·
Reassessing copper binding in lignin inspired cluster models
Reassesses copper–lignin cluster calculations and the source lineage of earlier multivalency interpretations.
Démarche
Repository and original ORCA output checks reproduce selected electronic binding-energy calculations and compare their consistency.
Portée et questions ouvertes
Electronic cluster energies, solution association constants and wash retention are different observables. The calculation does not establish practical retention performance.
Prépublication · Zenodo ·
Finite window diagnostics and noise channels in a kicked Ising chain
Distinguishes dynamical responses that receive the same finite-time return label in a kicked Ising chain.
Démarche
A reproducible finite-spin model compares signed odd-period and Fourier diagnostics under specified perturbations and noise.
Portée et questions ouvertes
Finite-window stability labels are diagnostic-dependent; the model benchmark does not establish a universal dynamical phase.
Prépublication · Zenodo ·
Reassessing constant claims in EEG band boundaries and their source provenance
Audits claims connecting EEG band-boundary estimates to selected mathematical constants.
Démarche
Published numerical summaries, estimator eligibility and code provenance are checked without reprocessing raw EEG.
Portée et questions ouvertes
Different estimators retained different participant subsets. The nearest-constant comparison depends on the estimator and aggregation.
Prépublication · Zenodo ·
Ratio dependent response curves and the limits of quantum biological transduction claims
Reassesses whether ratio-dependent response curves identify a quantum-biological transduction mechanism.
Démarche
Nondimensionalization, explicit error bounds and classical comparisons separate curve shape from absolute rate dependence.
Portée et questions ouvertes
A broad response peak is not mechanism-specific, and a fixed rate ratio need not preserve the absolute response.
Prépublication · Zenodo ·
A descriptive reanalysis of prompt context differences in the Trinity score corpus
Describes prompt-context score differences in the deposited Trinity corpus.
Démarche
A numerical audit compares the finite score table, model-label weighting and shared-label subsets.
Portée et questions ouvertes
The deposited scores are not independent trials. Context-associated differences do not by themselves establish a general property of models.
Prépublication · Zenodo ·
Measurement and source limitations in shared constraint embedding experiments
Reassesses evidence previously described as co-reactivity to shared constraints.
Démarche
Pinned sources, synthetic generators, historical outputs and diagnostic definitions are reconciled.
Portée et questions ouvertes
Planted lattice structure, deterministic replay and changing multiple factors together limit the original emergence interpretation.
Prépublication · Zenodo ·
Model and readout dependence of radical pair yield anisotropy
Studies how radical-pair yield anisotropy depends on model conventions, reaction operators and readout.
Démarche
Historical one-nucleus conventions are separated from a newly frozen finite computational benchmark.
Portée et questions ouvertes
A finite observation-window maximum can be absent from the complete-yield scan. Directional contrast depends on the specified reference and noise model.
Prépublication · Zenodo ·
Complete-tail projections and finite multiscale experiments for a discrete Riemann hypothesis criterion
Develops complete-tail projection identities and finite multiscale diagnostics for a discrete Riemann-hypothesis criterion.
Démarche
Weighted-Hilbert-space approximants retain the infinite tail through scaling, projection, transport and Mellin representations.
Portée et questions ouvertes
The reported computational campaigns are finite. An all-scale approximation or positivity obligation remains necessary.
Manuscrit de recherche · Zenodo ·
Mode-Selective CuI Anchoring on Lignin-Inspired Fiber Interfaces: A DFT Cluster-Model Study
Provides a computational audit pipeline for CuI anchoring on lignin-inspired molecular interfaces.
Démarche
DFT cluster calculations compare selected copper environments and molecular proxies for lignin and related interfaces.
Portée et questions ouvertes
Cluster-model energetics do not directly establish solution binding constants, wash durability or material performance.
Prépublication · Research Square ·
The Stress Paradox: Cardiovascular State Modulates Cardiac Radiofrequency Absorption While Modern Buildings Attenuate Natural Electromagnetic References
A deposited preprint examining cardiovascular state, radiofrequency absorption and attenuation of natural electromagnetic references by buildings.
Portée et questions ouvertes
Read the deposited manuscript for its assumptions, methods and evidence. This entry retains its preprint status.
Brouillons en cours
Ce sont des manuscrits exploratoires inachevés. Leurs questions ouvertes et leurs limites restent explicites.
Version de travail — inachevée · Zenodo ·
Systems pharmacology identifies dermal papilla–specific transcriptomic reversal of androgenetic alopecia by lithium–zinc pathway convergence
Explores transcriptomic reversal hypotheses for androgenetic alopecia using computational pathway comparisons.
Portée et questions ouvertes
Les substituts pharmacologiques, les contrôles de combinaison et les statistiques nécessitent des vérifications supplémentaires ; il ne s’agit pas de preuves cliniques.
Version de travail — inachevée · Zenodo ·
Gravity as Emergent Coherence
Explores a conjectural relationship between coherence and gravity.
Portée et questions ouvertes
Le mécanisme gravitationnel est conjectural ; l’analyse des résonances reste à vérifier.
Version de travail — inachevée · Zenodo ·
Scale-Free Dynamics, Bridge-State Control, and a Seven-State Neural Architecture: Evidence from Cross-Frequency EEG Analysis
Explores cross-frequency EEG state structure and a proposed bridge-state architecture.
Portée et questions ouvertes
Ce travail recoupe la version antérieure ; les définitions des états et les résultats à l’appui nécessitent des vérifications supplémentaires.
Lire l’original · DOI 10.5281/zenodo.18560751
Zenodo · 10.5281/zenodo.18477273 · Version de travail — inachevée
Version de travail — inachevée · Zenodo ·
The Vertical Ladder: Observer-Invariance as a Criterion for Distinguishing Intrinsic from Projected Structure
Proposes observer invariance as a way to distinguish intrinsic structure from projected patterns.
Portée et questions ouvertes
L’indépendance des méthodes et les exemples à l’appui doivent être mis en cohérence.
Version de travail — inachevée · Zenodo ·
Emergent Discreteness from Stability Boundaries in Reversible Dynamics: Cross-Domain Numerical Validation Suite
Explores whether stability boundaries can generate discrete structures in reversible dynamical models.
Portée et questions ouvertes
Les seuils de détection et l’affirmation d’universalité proposée nécessitent des vérifications supplémentaires.
Version de travail — inachevée · Zenodo ·
PhiGrid: An Open Platform for Studying Emergence Patterns in Transformer Networks with Persistent Memory
Proposes a software platform for investigating patterns in transformer networks with persistent memory.
Portée et questions ouvertes
La proposition de plateforme, le cadre mathématique et les mesures de calibration nécessitent davantage de développement.
Version de travail — inachevée · Zenodo ·
The φ-Framework: Golden Ratio Scaling Across Physical Domains
Explores a proposed golden-ratio scaling framework across several physical domains.
Portée et questions ouvertes
Les affirmations dans différents domaines nécessitent des vérifications distinctes des sources et des comparaisons plus solides.
Les titres enregistrés et les notes détaillées suivent les sources originales en anglais.