Our Work

Describiremos un modelo matemático que se podría utilizar para describir cada uno de los picos de un espectro de Absorción vs. la Longitud de Onda (una función para cada pi…

Describiremos un modelo matemático que se podría utilizar para describir cada uno de los picos de un espectro de Absorción vs. la Longitud de Onda (una función para cada pico).

Frente ao avanço da plataformização curricular padronizada no ensino básico, ao capitalismo de vigilância e aos riscos da semiformação (Halbbildung…

Frente ao avanço da plataformização curricular padronizada no ensino básico, ao capitalismo de vigilância e aos riscos da semiformação (Halbbildung), este artigo investiga como o diálogo entre a Educação Ambiental Crítica, as Tecnologias Sociais e as Metodologias Ativas de Aprendizagem pode fomentar uma práxis educativa emancipatória e sustentável. A partir de uma abordagem qualitativa e documental fundamentada no Materialismo Histórico-Dialético e na Pedagogia Histórico-Crítica, realiza-se uma síntese teórico-empírica comparativa entre o modelo didático corporativizado em slides da rede estadual paulista, os instrumentos pedagógicos de investigação científica da Pedagogia da Alternância na Escola Agrícola Terra Nova (MT) e as metodologias ativas nos Projetos Integradores da Univesp. Os resultados demonstraram que, enquanto os materiais digitais padronizados promovem uma Educação Ambiental periférica, funcional e despolitizada (hegemonia de 82,0% da corrente sistêmica e 76,9% da científica), expropriando a autoria docente e reduzindo o aprendizado à reprodução e mercantilização de dados comportamentais, as tecnologias sociais baseadas na alternância (Plano de Estudo e Caderno de Campo) e a orientação virtual de intervenções ativas resgatam a densidade dos conteúdos científicos, a experiência profunda (Erfahrung) e o trabalho como princípio educativo.

This is AI-generated research: the results, proofs and code were found and written by AI. Credit goes to all the humans whose work it builds on.We consider online linear regression in dimension one wh…

This is AI-generated research: the results, proofs and code were found and written by AI. Credit goes to all the humans whose work it builds on.We consider online linear regression in dimension one when the features are revealed one at a time, with no bound known in advance, and the regret is measured against every linear predictor, with no bound on its parameter. For outcomes bounded by $B$ and a horizon $T\ge355713$, both known to the learner, the minimax regret lies between $B^2(3\log T-2\log\log T-15.2)$ and $B^2(3\log T+0.37)$: an explicit learner attains the upper bound, and an explicit randomized adversary, with features in $[0,1]$, forces the lower bound. A learner that knows neither $B$ nor $T$ has regret at most $M^2(3\log T+4\log\log(T+2)+12+2/\sqrt T+4/T)$ on every play, with $M$ the largest absolute outcome. Lean 4 proves these bounds. Chen, Qian, Rakhlin and Zhivotovskiy (COLT 2026) showed that this regret is of order $\log T$, answering a question of Gaillard et al. (ALT 2019); we find its constant, $3$, which is $1$ with the features known in advance.The order $\log T$ is Theorem 5 and Proposition 6 of Chen, Qian, Rakhlin and Zhivotovskiy, Self-Normalized Martingales and Uniform Regret Bounds for Linear Regression (COLT 2026, PMLR 336; arXiv:2605.01628). The question is in Section 4.1 of Gaillard et al., Uniform regret bounds over $\mathbb{R}^d$ for the sequential linear regression problem with the square loss (ALT 2019, PMLR 98; arXiv:1805.11386).Lean 4 with Mathlib proves the main theorem with its constants, the bounds of the explicit learner and of the explicit adversary with every numerical step and the drift certificate, the extensions to bounded features and randomized learners, and the bound of the learner that knows neither $B$ nor $T$ with its corollaries, with no hypothesis and only the standard axioms.Code, Lean proofs and certificates: github.com/mt0-svg/regret-kappa

An influential paper by Worobey et al. and 17 other authors published in Science in 2022 claimed that case locations and nucleic acid samples showed that SARS-CoV-2 had first been introduced to humans…

An influential paper by Worobey et al. and 17 other authors published in Science in 2022 claimed that case locations and nucleic acid samples showed that SARS-CoV-2 had first been introduced to humans at the Huanan Seafood Market. Several lines of evidence suggest the introduction to humans was earlier and elsewhere.URL: https://econjwatch.org/file_download/1442/WeissmanSept2026.pdf?mimetype=pdf

This replication package accompanies the manuscript "Actionable Meta-Requirements for AI Robustness Auditing." It contains the materials used for the systematic literature review, literature-base…

This replication package accompanies the manuscript "Actionable Meta-Requirements for AI Robustness Auditing." It contains the materials used for the systematic literature review, literature-based meta-requirement derivation, practitioner survey, quantitative survey analysis, qualitative coding of practitioner comments, and final framework refinement. A README file describes the folder structure and the role of the main files.

Purpose – To give an overview of topics related to aviation and climate change. --- Methodology – A presentation of results from literature reviews and own calculations. --- Findings &ndas…

Purpose – To give an overview of topics related to aviation and climate change. --- Methodology – A presentation of results from literature reviews and own calculations. --- Findings – Renewable energy will by far not be sufficient to maintain flying at the level we know today (2019). It is therefore paramount to reduce flying as it happened already during the Corona pandemic. If we do not gently start changing the aviation industry now, laws of physics will tell us, with the consequence that our children will later struggle harder to make the necessary transformation.

Cette revue systématique et bibliométrique analyse le lien entre contrôle interne et risque opérationnel dans le secteur bancaire. À partir de 364 documents extraits …

Cette revue systématique et bibliométrique analyse le lien entre contrôle interne et risque opérationnel dans le secteur bancaire. À partir de 364 documents extraits de Scopus, le protocole PRISMA 2020 a conduit à sélectionner 42 articles, soumis à une analyse bibliométrique (VOSviewer) et thématique. Les résultats révèlent un champ en expansion mais fragmenté, dans lequel la littérature récente met en évidence des besoins de complément des référentiels classiques (COSO, Bâle) face aux transformations technologiques, comportementales et organisationnelles du secteur. Sur cette base, l'étude propose un cadre conceptuel exploratoire, le Contrôle Interne 4.0, articulant technologie intelligente, éthique opérationnelle, résilience dynamique et gouvernance intégrée, décliné en propositions de recherche testables, et débouche sur un agenda de recherche pour les économies émergentes Mots clés : contrôle interne ; risque opérationnel ; secteur bancaire ; revue systématique ; Contrôle Interne 4.0

Este modelo matemático diseñado, puede ser usado como Método Numérico para poder extrapolar y conocer la concentración inicial de un analito, cuando este decae en el…

Este modelo matemático diseñado, puede ser usado como Método Numérico para poder extrapolar y conocer la concentración inicial de un analito, cuando este decae en el tiempo (haciendo cálculos a partir de medidas posteriores al día cero que es la fecha de recolección de un fruto).

The conventional seesaw mechanism for neutrino masses relies on unobserved heavy particles and a high scale of new physics. We propose an alternative origin for the neutrino mass scale, derived direct…

The conventional seesaw mechanism for neutrino masses relies on unobserved heavy particles and a high scale of new physics. We propose an alternative origin for the neutrino mass scale, derived directly from the local geometry of the SU(2) gauge field. The standard Wilson action, by taking the trace of the SU(2) plaquette, retains only the gauge-invariant component c_0=1/2 Tr[P]. The remaining three components c_k (the Pauli coefficients) have a vanishing ensemble average by symmetry: ⟨c_k⟩=0. This vanishing average has been interpreted as indicating that these components have no physical effect. The Lamb shift shows this inference is invalid. The vacuum electromagnetic field has zero mean, yet the variance of its fluctuations shifts atomic energy levels measurably.We propose a reevaluation: the c_k are components of the field strength tensor: F_μν (x)∝∑_(k=1)^3 c_k  (x) σ_k. Their variance is therefore physical energy density: ∑_(k=1)^3 Var (c_k )=⟨Tr[F_μν F^μν ]⟩∝E_YM. This is a field-theoretic identity. Energy density creates mass. The seesaw mechanism was introduced to account for a mass that already resides in the variance, obscured because its mean vanishes, but physically present because variance constitutes energy.Quantum simulations with random sampling of the SU(2) gauge group confirm this mechanism. We ran planar lattices of 9, 16, and 25 qubits and cubic lattices of 8 and 27 qubits. The ensemble mean cancels (⟨Z^(⊗9)⟩≈0), while the fluctuation remains fully in place (Var(Z^(⊗9) )≈1). Reading the Pauli coefficients directly from plaquettes built from Haar random links, ⟨c_k⟩≈0 while Σ Var(c_k) = 0.751 ± 0.002, the exact value 3/4 fixed by the geometry of the SU(2) manifold, on every lattice size and topology; the same values are recovered on qubits with a Hadamard test. A neutrino probe qubit entangled with the gauge register shows exactly zero variance with the field off, and a variance that grows with Σ Var(c_k) with the field on. A companion classical Wilson campaign at L = 16, 24, and 32 finds the same signature under thermal dynamics: the site-level mass distribution has a lower edge that does not move when the volume grows eight-fold. The effective mass tracks the energy density of the gauge sector directly. Across three independent injection protocols, the mass collapses onto a single function of the measured variance. It relaxes to the geometric floor as the variance channel saturates. In the ground state, loops pierced by a center vortex carry a phase locked at π, W_odd/W_even=-1.00, a topological quantum lock that forms at the same physical size at four lattice couplings. We propose that this phase is a topological quantum lock held dynamically by the circulation of the field, distinct from the static reading of center vortices as kinematic defects that confine. Four dimensional simulations support it: one and three turns of a probe fix the held phase at exactly π at every loop size, the lock forms at one physical size, 1.05/√σ within 5 percent, across four couplings, and it does not depend on the gauge copy. The three twist modes of the variance settle at populations 0.692, 0.266 and 0.041, fixed by the group alone and matching the measured electron neutrino row 0.677, 0.301 and 0.022 with no adjustable parameter. The flavor composition of the neutrino thus follows the orientation of the SU(2) variance, and the mixing angles become outputs of the group geometry rather than inputs. This is the first link between the variance of a gauge field and the observed mixing of the neutrino. The electron neutrino inherits this division because it is the flavor whose charged current coupling to SU(2) is active in ordinary matter, while the muon and tau neutrinos share the remainder nearly evenly, as measured. This principle unifies mass generation in the electroweak and strong sectors, suggesting a common geometric origin for hadron and neutrino masses without requiring new particles or high energy scales.