Our Work

Die hochschulpolitische Diskussion um Studienerfolg, Studienabbruch und das Studierendenerleben gehört seit Jahren zu den prominenten Themen der empirischen Hochschulforschung (Heublein et al., 2…

Die hochschulpolitische Diskussion um Studienerfolg, Studienabbruch und das Studierendenerleben gehört seit Jahren zu den prominenten Themen der empirischen Hochschulforschung (Heublein et al., 2017; Pelz et al., 2024). Parallel hat die Digitalisierung der Lehre, beschleunigt durch die SARS-CoV-2-Pandemie, neue Forschungsfragen aufgeworfen - etwa zur studentischen Partizipation in digitalen Formaten (Lohberger & Braun, 2024) oder zur Eignung digitaler Unterstützungsangebote für heterogene Studierendengruppen (Sielschott et al., 2025). Künstliche Intelligenz (KI) ist in diesem Zusammenhang in doppelter Hinsicht relevant: als Forschungsgegenstand - wie Studierende KI in ihrem Studienalltag wahrnehmen (Hüsch et al., 2025; Marczuk et al., 2025) - und als Forschungswerkzeug, das die Auswertung großer qualitativer Datenbestände potenziell verändert (Than et al., 2025). Die qualitative Inhaltsanalyse nach Mayring (2022) ist ein methodisches Rückgrat der empirischen Hochschulforschung: Sie verbindet regelgeleitete Kategorienbildung mit Intercoder-Reliabilitätsprüfungen und ermöglicht die systematische Auswertung offener Antworten, Interviewdaten und Dokumente. Mit dem Aufstieg generativer Sprachmodelle stellt sich die Frage, ob und in welcher Form große Sprachmodelle die menschliche Zweitkodierung methodisch sinnvoll ergänzen können. Internationale Vorarbeiten zeigen, dass solche Modelle in bestimmten Annotationsaufgaben menschliche Plattformarbeitende hinsichtlich Übereinstimmungsmaßen erreichen oder übertreffen (Gilardi et al., 2023; Ziems et al., 2024). Andere Studien dokumentieren systematische Verzerrungen, Halluzinationen und Reliabilitätsschwankungen je nach Modell, Prompt und Domäne (Misra et al., 2026; Pangakis et al., 2023; Reiss, 2023; Törnberg, 2023). Vor diesem Hintergrund leistet der vorliegende Beitrag einen methodischen Beitrag zur qualitativen Hochschulforschung: Er untersucht an einem konkreten Anwendungsfall, unter welchen Bedingungen ein großes Sprachmodell (Claude Opus 4.7) die menschliche Zweitkodierung in Mayring-basierten Inhaltsanalysen reliabel ergänzen kann. Anhand einer dreifachen Parallelkodierung - Erstautorin, menschlicher Zweitkodierer und LLM - wird die Intercoder-Reliabilität systematisch ausgewiesen und durch einen formalen Bootstrap-Test der Kappa-Differenz statistisch abgesichert. Als Anwendungsfall dient eine Erhebung studentischer Vorstellungen von einer „perfekten Hochschule“ (N = 174) an einer Hochschule für angewandte Wissenschaften; die inhaltlichen Befunde dienen der methodischen Illustration. Der Beitrag versteht sich als Fallstudie ohne Anspruch auf Repräsentativität und expliziert seine Grenzen in Abschnitt 6.3.

The aim of the article is to systematize the historical evolution of the agribusiness category, distinguish it from the concepts of the agro-industrial complex and agricultural entrepreneurship, and s…

The aim of the article is to systematize the historical evolution of the agribusiness category, distinguish it from the concepts of the agro-industrial complex and agricultural entrepreneurship, and substantiate the scientific content of the "agri-industrial business" category in the context of modern value chains. The study employed methods of systemic analysis, comparative-categorical analysis, a historicallogical approach, and functional decomposition. As a result, agri-industrial business was interpreted as a category that unites resources and services, agricultural production, processing, logistics, marketing, and consumer relations into a single entrepreneurial system. The differences between vertical and contractual integration were revealed in terms of ownership, control, risk distribution, and market access for small businesses. Furthermore, a multi-level management mechanism operating at the enterprise, national, and international levels was proposed. The discussion results show that terminological updating alone is insufficient; this category must be reinforced with real institutions that coordinate economic interests, information, and risks across the value chain.

One cumulative programme, one canonical DOI. This edition reunites the original collection and all material previously issued in a separate continuation. The permanent project DOI is 10.5281/zenodo.22…

One cumulative programme, one canonical DOI. This edition reunites the original collection and all material previously issued in a separate continuation. The permanent project DOI is 10.5281/zenodo.22678085. All earlier mathematical files, complete LaTeX, human citations and source archives are retained. Selected main papers remain separately readable; the full set of 98 original PDFs is also preserved in one ZIP. This is a publication consolidation, not a new proof or a change to the mathematical scope. Split-Zero cohomology and the zeta-function research programme is a cumulative collection of mathematical papers, complete LaTeX proofs, formalization sources, calculations and literature-based research notes. Its central question is how the zeros of the Riemann zeta function can be studied through explicitly constructed arithmetic cohomology, spectral actions and quantitative metric estimates. The collection also retains the related heat-flow, arithmetic-trace, fluid, complexity and thermal-geometry investigations from which parts of this programme developed. Project and reading guide · Research contents · Machine-readable workbench index. Papers remain separate readable PDFs; the source archives supply the corresponding editable mathematics and supporting materials. What the collection contains Split-Zero coefficients, geometry and cohomology. The construction retains a ring's internal zero and a separate symbol for absence, so a zero element in a specified support fibre is not confused with an absent fibre. The papers construct coefficient maps, quotients, homology, finite spectral jets and their relation to the infinite theta quotient. Deligne's Weil II provides a motivating source for the cohomology–duality–weight-control programme. Its direct-image theorem 3.3.1 is stated for schemes of finite type over the integers; its image-purity corollary 3.3.6 is over a finite field. Neither is asserted to apply automatically to the Riemann zeta function: the actual coefficient objects and their comparison maps must be constructed. Read the cohomology and weight-control paper and the foundational objects and maps. Full-prime cohomology and complete support. The 75-page full-prime paper develops relative principal parts, the actual degree-p operator and a local complex weight-one identity from the full arithmetic prime space. The complete-support paper recovers mixed information lost by separate branches. The mixed-Weil paper applies the actual prime operator, includes the full Archimedean contribution and proves positivity for its specified three-pulse family, while computing a distinct negative signed joint trace and the original companion object's nonzero distributional boundary. These results do not establish global interior-cohomology purity or decide RH. Full-lattice geometry, graph observations, heat and residue duality. The 208-page reconstruction carries full support through prime cohomology, finite jets, exact graph metrics and prime-pulse observations with their complete costs. The 103-page Connes–Consani component develops labelled arithmetic sites, tropical spectra, sheaf, divisor, twistor and tensor maps. The 40-page heat-endpoint paper calculates the original support and heat maps, cluster-energy expansions and holonomy receivers. The 157-page supported-zero and residue collection retains its earlier prime, heat and class-field calculations and adds the intrinsic supported-zero identity and exact Jacobian map from residue duality to Weil trace and cotangent cohomology. Complete proof sources retain all nilpotent and endpoint terms; exploratory companion calculations carry their own review status. These papers do not claim an RH decision. Integral collision prisms, extension classes and original heat holonomy. The 177-page supported-zero and residue collection retains its preceding21 proofs and adds the complete integral extension group with both connecting maps, the bounded collision prism with its divisor quotient, and every supported prime pullback. The separate 60-page heat-collision and holonomy paper retains the40-page endpoint argument and calculates the full conductor, cotangent linking form, complete-unit connection, monodromy, reflected-trace defect, arithmetic commutator and joint two-prime fixed-support action. The original units, coefficients, nilpotents and support labels remain. Neither addition claims an RH decision. Integral heat monodromy and supported trace responses. The 15-page supplement proves regular monodromy on the original integral heat-collision algebra at every actual multiplicity, evaluates the first and next trace responses of every coefficient layer, and carries the connection residue and parity projections through the complete support lattice. Its eight-state application retains the length-four collision algebra and calculates the exact multiplication defect and prime-spectrum map. The endpoint trace and remaining global arithmetic sign question are not replaced by these time-response calculations. Original heat-Cauchy arithmetic, reflection index and fixed tests. The 210-page supported-zero, prime, heat and residue collection retains its preceding 24 proofs and adds six complete derivations: the actual zero-distribution heat response, all Cauchy/Gamma/prime and endpoint terms, the full mixed-prime derivative recurrence, the Cauchy–Weil matrix criterion, the exact reflection index of the actual zero pairing, and the finite-jet/exterior decomposition with higher heat coefficients. The index counts actual exchanged zero pairs without asserting that any exist. The complete positive-matrix condition is equivalent to RH; positivity of that entire family is not established. The separate five-page fixed arithmetic test paper computes the first and next heat-trace coefficients, the original Cauchy frame and its inverse, and every endpoint jet of fixed vanishing tests. All original multiplicities and support coordinates remain. Boundary contact, residue data and arithmetic trace. The 16-page paper computes the full analytic-unit contact equation for the supported reciprocal and its exact map into the original arithmetic trace derivative. It identifies the sharp residue orders needed by fixed-test and determinant corrections, retaining the original time and all higher test jets. Complete theta integrals with directed intervals and explicit quadrature and tail bounds reproduce the supplied nineteen-coefficient rational witness: negative at heat time −2, positive at time 0. This is not a time-zero negative witness or a certification of the companion manuscript's separate 32-dimensional inertia. The full supplied sources, complete proofs, human citations and corrected retained LaTeX accompany the paper. Primitive tests and prime-window arithmetic. The 235-page supported-zero, prime, heat and residue reader retains 34 complete derivations. Its new calculations give the von Mangoldt operator from divisibility projections with both support corrections, the explicit 55/192 short-support bound, a fixed primitive test detecting every possible off-critical zero, and the exact prime-window formula with a positive Archimedean remainder. The uniform bound equivalent to RH is not proved. The classical bump and its product zero multiplicities are credited beside the construction to Arias de Reyna, with the full coordinate comparison. All complete LaTeX, proof texts, citations and supplementary sources are retained. Original zeta and faithful theta-source return. The eight-page correction works from the integer-dilation operator whose Mellin multiplier is the original zeta function. It proves the Gaussian theta return with both endpoint moments, calculates two boundary coordinates per prime and a faithful augmented receiver, and retains all exceptional local jets, heat drift and potential terms, and original metric factors. The original seminorm still has the calculated prime-boundary kernel. This is a completed correction of the stated comparisons, not a claim that every preceding estimate has been reconstructed or that RH has been decided. Full original-zeta receiving calculations and faithful integer heat observations. The 336-page, 47-proof collection retains its complete supported-zero, prime, heat, residue, integral, geometric and spectral calculations and reconstructs their theta, meromorphic heat, compact-Weil and signed-Cauchy receiving maps from the original zeta function. Its Gamma factors, endpoint terms, trivial zeros, pole, local jets and support labels remain explicit. The raw signed and compensated pairings retain their exact correction; RH remains unresolved. A separate seven-page sampling addendum proves an explicit integer-sample inverse for the original Gaussian source, transports the supported-zero mass and every label, and proves the precise information lost by circle periodization and the instability of half-integer sampling. Arithmetic theta sources and analytic control. The central analytic complex is [V → B], with theta map Θ and quotient B/ΘV. Scaling acts by D = −x∂x. The underlying integer-dilation operator has Mellin multiplier ζ(s). The chosen differential Gaussian theta source has the full multiplier s(s−1)π^(−s/2)Γ(s/2)ζ(s), written 2ξ(s) in earlier editions. The current source reconstruction retains every factor, both endpoint moments and the exceptional local modules; older proofs retain their original scopes until their dependent estimates have been reconstructed. The papers carry finite zero packets, including multiplicities, through the original source maps and least-norm quotient metrics. They develop source recovery, residue and period maps, tensor and exterior constructions, Mellin and Gamma estimates, orthogonal-polynomial kernels, Hankel–Toda determinant identities, boundary terms, and growing-degree volume and covariance calculations. The cumulative Split-Zero reader, its proof companion, and the separate kernel, conductor and source-transport papers retain the complete arguments. Conductor geometry and spectral observation. Further papers calculate the original relation and conductor maps, kernels, quotient volumes, real-period resonances, inverse spectra and finite observation bounds. The native spectral paper, real-period paper, resonance-direction paper and inverse-power observation paper have distinct, explicit scopes. The local-branch paper constructs the coupled spectral module; the covariance paper calculates the arithmetic-kernel return; and the activation and current paper evaluates source activation and exact projected-current formulas. The original-observation paper evaluates fixed-coefficient inverse-power subquotient determinants, measured heat and conductor-detected eigenlines in the original metric. The actual-Xi and uniform inverse-rate paper calculates the Xi-jet pullback, a local arithmetic interval certificate and the two original geometric inverse rates. Geometric examples are not represented as identified arithmetic zeta-zero packets. The current, source-overlap and collision paper calculates the sharp joint phase region, measured-mass interval and actual source-step receiver, and bounds the full arithmetic collision resolvent and determinant on its exterior frequency region. The actual terminal phase, marked sign and interior contribution remain unevaluated. The real word currents and source-step paper proves constant polynomial-current inertia throughout actual adjacent-source additions, observed sign dimensions and the complete signed-product formula, and calculates the prescribed current from endpoint traces and weighted collision data. Its original terminal imaginary pairing remains unevaluated; the preceding response-coefficient correction remains active. The support-transport paper calculates both measured support directions, ordered mixed-current recovery, observed collision, metric-stable source steps and controlled evolution. It explicitly corrects the response-coefficient identification in edition 027 while preserving its sealed files. The terminal-response paper evaluates the two original boundary-denominator amplification rates separately, controls the phased current-sign amplitudes of the observed terminal class and bounds its diagonal current correction, while retaining the unresolved imaginary cross-coordinate on the exact conductor domain. The current and effective-cutoff paper connects the original kernel and current profiles, evaluates opposite current signs on the actual observed plane and the intrinsic evolution asymmetry, and imports the established conductor norm comparison to bound the separately marked terminal current while retaining its unresolved residual sign. The complete-cutoff profile paper evaluates the original kernel spectrum and volume throughout the source window, carries that profile through actual increments and complete fixed subquotients, and calculates arbitrary-cutoff measured phase and sharper original heat estimates with the full angle defect. The kernel-coefficient and measured-frequency paper evaluates the original four-cutoff kernel determinant on the retained fixed-period simple-quartet domain, identifies its elliptic constant, and controls the complete measured determinant and phase while preserving the unfinished status of individual arithmetic-current signs. The arithmetic-probe and kernel-mass paper decomposes the original kernel determinant into positive observation layers, sharpens the interval obtained from the same mass samples, and calculates the terminal correction between hidden-coordinate and full-ambient minima while retaining canonical projective sections and complete currents. The original-observation recovery paper proves the zero-frequency kernel-angle spectral map, sharp scalar information loss and leading-coefficient recovery, with a bounded-power remainder estimate and original pole and current formulas. The original-growth and resolvent paper controls complete-row sampling and root errors, reconstructs the full complex kernel correction from the measured action, bounds complete-source metric errors and complementary growth, and evaluates the separate four-label receiver through its first singular corrections. The original-angle and full-word paper controls every original projection direction, bounds directed spectral movement over the complete cutoff window, identifies phased source innovations and extracts the determinant using actual source columns. The invariant-determinant and collision paper evaluates an actual conductor word image, retains the collision graph metric, calculates the complete finite invariant determinant and splits positive cutoff losses into source, observed and kernel energies. The moving rational-family paper controls the original kernel energy throughout activation, evaluates late observed memory, and carries moving rational filters and arithmetic collisions through complete primary sectors and corrected chain maps. The canonical memory and polynomial-filter paper controls the complete memory including zero regularizers, evaluates original energy changes, and retains colliding filters, repeated-pole observation masses and the slow energy-graph spectrum. The critical inverse-activation paper evaluates the critical inverse-power sector, keeps every singular direction in the original observed arithmetic action, and transfers these bounds to the original kernel and complementary determinants at their stated source status. The source-family and observed-heat paper controls the actual two-column update, changing positive source weights, rational detection and the complete observed-memory contribution. The inverse-sector quotient paper gives exact quotient and shifted-source maps, bounds the original covariance-angle change and carries heat and current formulas through that reduction. Formalization and reproducible calculation. The Lean sources and formalization notes cover specified algebraic, quotient and homological interfaces. Source packages retain exact finite checkers, numerical or interval calculations where supplied, figures, manifests and verification records. A Lean check of an interface, a finite algebra test and an analytic proof are documented as different forms of evidence; none silently certifies the others. The wider zeta literature and connected investigations. The separate zeta reader, fluid/Navier–Stokes reader, de Bruijn–Newman heat-flow reader, Connes/arithmetic-trace reader, geometric-complexity reader and vacuum and thermal-geometry reader preserve the earlier research routes and their stated limitations. Related Unified Flow calculations concern specified black-hole, thermal, spectral-zeta and geometric-flow probes. The CUE research map connects existing random-unitary-matrix, Mellin and finite-phase-space work. An analogy or proposed connection is not presented as an established implication between major conjectures. How to read and assess the work This is an ongoing research programme, not a claimed proof or disproof of RH or the other major conjectures discussed in the connected notes. The remaining arithmetic estimates and identifications are stated in the individual papers. Original coordinates, masses, multiplicities, signs, relation spaces and metric factors are retained. Earlier editions remain available because later arguments depend on their complete constructions; an older paper's use of "current" refers to its own edition. Human authorship and mathematical provenance are part of the work, not optional packaging. The papers and source ledgers identify primary human sources, precise versions and theorem or formula locations, including Deligne, the named authors of NIST DLMF chapters, Akemann–Vernizzi and Johansson where their results are used. Programme citations link to the actual public proofs; a public reconstruction is identified as such and is not presented as the original private correspondence. Per-file rights and existing attribution remain in force. Papers incorporated from the former continuationFull-support cohomology and graph/prime-pulse reconstruction (208 pages); complete LaTeX.Original-zeta sectorial return, augmented inverse and full arithmetic boundary (385 pages); complete LaTeX.Original zeta, full support and arithmetic trace (37 pages); complete LaTeX.Localized adelic coefficients on the full support base (21 pages); complete LaTeX.Original endpoint resonance, heat range and prime class (10 pages); complete LaTeX.Original zeta, adelic correspondences and full Hurwitz boundary (34 pages); complete LaTeX.An evaluated arithmetic pairing for the original endpoint test (7 pages); complete LaTeX.The 385-page collection retains 52 complete main proofs and gives ten full original-zeta receiving calculations. The separate 34-page reader corrects the scalar theta-origin, transported product and unit, and embedded-fibre account, then constructs the actual adelic and full-character maps. The 37-,21- and10-page papers retain their distinct proofs and scopes. The seven-page pairing successor evaluates the formerly open scalar at t=1/32: 50<K<64, with the complete original theta residual, all prime powers and full Gamma contribution. The same test retains two negative linear boundary moments. This is a result for that specified test, not whole-space positivity or an RH decision.The complete source archive includes every programme LaTeX source and dependency for these PDFs, the separate source-only coefficient correction, figures, checkers, owner verification scopes, pinned proof links and the exact 355-page predecessor. The machine-readable proof index supplies clickable external source locations. Human authorship and source-use ledgers are retained; protected primary archives are linked to their original hosts. Publication adopts the recorded owner reviews at their stated scopes and is not a new independent proof audit of all pages.Source recovery and mixed-boundary continuationThe 24-page original-zeta, faithful source-recovery and mixed-boundary paper computes the actual radial adelic source, every prime-boundary coordinate, a faithful joint inverse, higher Koszul connecting classes and both endpoint characters. For finite rational mixed profiles it gives an exact finite test of the entire boundary, with an actual adelic antecedent, compact characters and all initial-term/support corrections retained. These statements do not assert positivity for every Weil test or decide RH.Complete LaTeX; complete source addition; preceding pinned proof index. All nine previous continuation files remain unchanged and separately downloadable; the preceding edition has eight separate papers, including the cohomology preview. Both complete source archives are retained. The original base collection and all its files remain unchanged. The correction to the new reader concerns only the statement that human primary works are linked at their original hosts rather than included as archives; its mathematical proof bodies are unchanged.Original-zeta detection and supported-zero source-plane cohomologyThe 403-page collection retains 54 full proofs. Its resonant cosine test detects every original nontrivial zero at the fixed time 1/32. The paired-source calculation gives exact transport, a full arithmetic Weil matrix and a global translation criterion; the bound required by that criterion is not proved. The original small-time factor is retained even where all its time derivatives vanish at zero. The earlier 385-page edition remains an unchanged separate download.The 16-page source-plane paper evaluates the original cosine/sine Weil form, proves its arithmetic prime-boundary map, and computes supported-zero multiplication in that original metric. Its square-zero metric defect has a calculated vanishing time, local cohomology and full integer-prime module support. Coefficient heat time and theta time remain distinct. A general-ring proof sentence is corrected by the meet-prime label argument; no theorem, metric or numerical certificate is changed. These are finite-source and exact-map results, not universal positivity or an RH decision.403-page collection: complete LaTeX.16-page source-plane paper: complete LaTeX.Complete source addition and preceding pinned proof index. The preceding edition retained all twelve earlier downloads and added four files, giving ten separate PDFs and three complete source archives; all sixteen remain in this edition. The cohomology preview, stable whole-project identity, human provenance, source rights and original base collection are unchanged. Mathematical-owner review and publisher source/byte checks keep their recorded scopes.All-prime tensor boundary weightsThe eight-page boundary-weight paper calculates all tensor coinvariants and every rational group-homology degree of the original arithmetic boundary. It retains the actual idele action, evaluation and integral factors, both sign projectors and the full support lattice. It exhibits a nonzero mixed tensor class erased by the two-endpoint tensor quotient and gives the exact connecting map into the original Schwartz-kernel homology. This supplies a boundary term for the continuing weight construction, not an interior purity theorem or an RH decision.Complete LaTeX; complete source addition; current pinned proof index. Result SZ-20260923-153 retains full proofs BW1–24 and independent TB1–29, citations to Deligne and Connes, complete diagram sources and 187 exact finite checks rerun from the package. Every final PDF page was inspected. All 16 previous downloads remain unchanged: this edition has 11 separate PDFs, 4 complete source ZIPs and 4 proof-index files. The 208-page cohomology preview, stable project title, creator credit, rights and base collection are preserved. Geometric weight control, positive quotients and original-character lifting This continuation integrates three collaborating strands of the Split-Zero cohomology and zeta-function research programme. It retains the programme's original support construction, theta and Mellin sources, arithmetic cohomology, spectral actions and exact comparison maps, alongside the preceding analytic estimates, formalized interfaces, connected investigations and research self-audit preserved in this collection and its linked base collection. Three separately readable cumulative readers accompany the complete editable source archive and a machine-readable result/dependency index. The geometric reader reconstructs Deligne's weight argument and develops the actual coefficient sheaf, nearby-cycle, Gysin, continuous-dual and geometric-cover maps. Its 63 source parts include the full adjoint-defect invariant-cycle comparison. The positive-quotient reader identifies the defect removed by positive compression, retains the complementary original Weil contribution and higher jets, and constructs the exact strong-dual Gysin receiver. It also classifies the jointly continuous positive Hermitian forms satisfying the original integer-transfer relation, calculates their common completion and proves that all such forms on the complementary specialization obstruction vanish without inferring that this obstruction itself vanishes. The original-character reader gives Mellin-character lifts, restriction extensions, residue pushouts and supported comparison triangles with the full Gamma germ and endpoint data retained. The papers distinguish proved internal calculations from their unproved application to arithmetic purity. In particular, vanishing of a quotient constructed to remove a defect does not establish vanishing of that defect in the original source. This edition claims neither a proof nor a disproof of the Riemann hypothesis or a new P-versus-NP implication. The human investigator supplied the originating construction, mathematical agenda and corrections; collaborating AI tasks developed and checked the recorded calculations. Human mathematical sources, including Deligne, Connes–Consani, Meyer and the classical analytic and functional-analytic inputs, receive explicit citations and source locators. The source archive includes the programme's LaTeX, Markdown, figures, reproducible figure scripts, proof index and exact review/build record. It excludes private transcripts and human-author archives for which redistribution permission has not been established. Earlier downloadable papers and sources remain unchanged, and the full-support cohomology paper remains the reader preview. Complete matching GitHub source edition · 020-geometric-weight-control-complete.pdf · 021-positive-quotient-and-gysin-complete.pdf · 022-original-character-lifting-complete.pdf · 023-cohomology-weight-and-character-complete-sources.zip · 024-cohomology-result-and-dependency-index.json Point-of-use human-source correction in the analytic cumulative papers The FRG2 Gamma-polynomial calculation now credits the classical Meixner–Pollaczek family at the formula's actual use, with the authored NIST DLMF chapter and its exact parameter dictionary. This corrects an omission: the source bibliography and HPS comparison were already present, but the local FRG2 paragraph named only project notes. No equation, argument, mass, phase or previous source credit is removed or changed. This addition contains complete rebuilt 837-page cumulative and 263-page supplement PDFs, all six complete corrected LaTeX sources (including later inherited copies), and a reversible citation-change record. Four source copies were not separately rebuilt as PDFs. Earlier editions and the cohomology, positive-quotient and character-lifting readers remain available. The stable project description, creators, rights and cohomology preview are retained; this correction neither replaces that work nor asserts a resolution of RH. Complete matching GitHub source edition · 025-frg2-cited-cumulative.pdf · 026-frg2-cited-proof-supplement.pdf · 027-frg2-complete-citation-sources.zip · 028-frg2-citation-changes.json Full-source tensor, original-divisor and arithmetic-return continuation This addition integrates the subsequent completed work of the Split Zero to GCT bridge, the Split-Zero cohomology calculation, and the identity–absorber research task. It contains full-source Gaussian spectral synthesis and completed tensor comparison; positive source-measure classification and original-zeta divisor identities; and the derived arithmetic return, full Gram harmonic comparison, all-dilation forcing identities and closed-support restriction cross. Three complete continuation PDFs accompany all editable LaTeX and Markdown proofs, reproducible programme figures, a machine-readable result/dependency index and an exact source-reading/review record. Earlier cumulative readers, structural constructions, analytic estimates, formalization, connected research and self-audit remain in this collection. The investigator's parityless support is distinct from recovered integer1. Original zeta factors, every zero multiplicity, actual source topologies and surviving cohomology are retained. Human sources are credited at use. This edition does not claim an RH proof, arithmetic purity or human peer review. The project title, complete preceding description, authorship, rights, all earlier files and the Split-Zero cohomology reading preview are preserved. The mixed contraction is explicitly scoped to its degree-one obstruction, with the surviving K0 recorded; an unfinished divisor follow-up is not presented as completed. Complete matching GitHub source edition | 029-full-source-synthesis-and-tensor.pdf | 030-positive-source-and-original-divisor.pdf | 031-arithmetic-return-and-closed-support.pdf | 032-complete-three-task-proof-sources.zip | 033-results-and-programme-relevance.json Source endpoint and residue-section continuation This addition extends the existing Split-Zero cohomology and original-zeta programme with a complete 647-page cumulative identity-absorption reader. It retains the preceding geometric, original-divisor and positive-form arguments and adds the full-dual Hardy observation, the meromorphic endpoint with its exact cover action, generation of the original vanishing-jet ideal by the full cover-discrepancy family, and exact comparison of residue sections and polynomial quotients. The earlier papers, analytic estimates, formalization, connected investigations and research self-audit remain part of this collection and its linked base collection. The complete LaTeX, Markdown proofs, reproducible figures and machine-readable result/relevance index accompany the PDF. Two additional full mathematical reviews, with editable TeX, prove the maximal common identity-induced quotient, compute finite-section resonance and joint uniqueness, and strengthen the regularized operator result to every positive Schatten exponent. These are separately identified companions, not claims that the historical 647-page body already contained those further results. Publication notes give the exact conjugation in a general receiver-coefficient formula; its omitted conjugation in an earlier zero-case sentence does not change that zero conclusion. The investigator's parityless support remains distinct from the recovered integer unit; complete-history reconstruction precedes coefficient arithmetic. Original zeta, Gamma and endpoint factors, full zero multiplicities, source topologies and raw cover degrees remain explicit. S. Waleed Noor's Hardy-space results and the inherited work of Deligne, Connes–Consani, Meyer and other human sources are cited at their uses, with reading coverage stated. These calculations do not establish arithmetic purity, geometric lifting vanishing or an RH proof. Independent AI reviews are not human peer review or a new Lean certification. All 647 rendered pages match the supplied reader after link repair. The source archive contains the full proof collection, not extracts; protected human-author archives and private correspondence are not redistributed. The project title, full preceding description, authorship, rights, all 33 preceding downloads and the full-support cohomology preview are preserved. The later GCT85 and complete-refinement packets are outside this specific frozen addition. Complete matching GitHub source edition | 034-source-endpoint-and-residue-section.pdf | 035-source-endpoint-complete-sources.zip | 036-source-endpoint-results.json Consolidated reading and source accessUse the reading guide and full file-location table and the machine-readable preservation index. All 100 original files and all 36 continuation files are represented without changing their contents; an identical preview PDF is stored only once as a separate download. The complete original 98-paper ZIP and both original complete-source ZIPs remain alongside all continuation sources. The separately staged 719-page successor is not included in this edition. The earlier collection descriptions above identify historical editions and their stated scopes.Added in this version (26 September 2026): an audit and reader by Claude. This version adds three files and removes none. 00-claude-opus-5-5-audit-reader.pdf (the default preview) is The split-zero programme around the Riemann zeta function: a reader of its results, with proofs (50 pages), an audit and reading of the programme's files carried out between 22 and 26 September 2026 by Claude (Anthropic), model claude-opus-5-5 (Opus 5.5), at maximum reasoning effort. Its purpose is to set out for a human reader what the programme has established, with full proofs of the central results and the exact location of every other proof; which routes are proved not to decide by themselves where the zeros of ζ lie, each within the exact scope of what was proved, without claiming that these routes have no bearing on the Riemann hypothesis; the connections with other fields and with the author's other workbenches; and the questions that remain open. The audit is not exhaustive, and the reader lists what was not audited. 00-claude-opus-5-5-audit-reader-latex-source.zip contains its LaTeX source, and 00-claude-opus-5-5-audit-notes-and-checks.zip the audit notes, the results register and the check scripts with their outputs; the same material is on the branch claude/claude-ab-grind-20260925 of the repository KokunoYumeto/zeta-function-research-reader. Nothing in this version proves or disproves the Riemann hypothesis. All earlier files and descriptions of this record are retained unchanged.Added in this version (26 September 2026, second Claude addition): version 2 of the Claude reader, and a reader of four further workbenches. The paragraph above describes the preceding version (10.5281/zenodo.22970703). This version replaces its three Claude files by six and keeps every other file unchanged. 00-claude-opus-5-5-audit-reader-v2.pdf (the default preview) is version 2 of The split-zero programme around the Riemann zeta function: a reader of its results, with proofs (58 pages). It applies a review of version 1 for understatement and missed generality as well as for overstatement, the verification of that review, and a referee pass on version 2 itself: four corrections, and stronger forms of more than twenty statements, among them the first Hurwitz jet on every sheet, a classification of the Hermitian forms compatible with the scaling transfer, and a Carleman criterion under which the cover discrepancies with degrees 2a3b are dense for every t > (log 2)(log 3)/(4π); its Appendix B lists every change. 00-claude-opus-5-5-workbench-reader.pdf is Four workbenches and where they meet (35 pages), a reader, with proofs, of the author's Erdős–Straus, Collatz, Erdős Problem 817 and Yang–Mills workbenches (with the Navier–Stokes, S⁶ and Jacobian-map material) and of their intersections, refereed three times, twice for understatement as well as overstatement. The two -latex-source.zip files contain the sources of the two readers; 00-claude-opus-5-5-audit-notes-and-checks-v2.zip contains the audit notes 00_–48_, the results register and every check script with its output, including the whole content of the version-1 archive. 00-claude-opus-5-5-audit-reader-v1-provenance.zip contains the three Claude files of the preceding version byte for byte (the version-1 reader, its source and the version-1 notes archive), with their MD5 checksums. The same material is on the branch claude/claude-ab-grind-20260925 of the repository KokunoYumeto/zeta-function-research-reader. Prepared by Claude (Anthropic), model claude-opus-5-5 (Opus 5.5), at maximum reasoning effort. Nothing in this version proves or disproves the Riemann hypothesis.

Nigeria's real-time payment infrastructure, built around NIBSS Instant Payments and now migrating toward the newer National Payment Stack, has quietly become one of the most heavily used pieces of fin…

Nigeria's real-time payment infrastructure, built around NIBSS Instant Payments and now migrating toward the newer National Payment Stack, has quietly become one of the most heavily used pieces of financial plumbing in the country, moving billions of naira a year through bank apps, fintech wallets, and point-of-sale terminals. What's less clear is what all that speed is actually doing for the micro-entrepreneurs who depend on it most: market traders, POS agents, and informal-sector operators who used to wait days for a bank transfer to clear and now often don't wait at all. This study examined the relationship between real-time payment infrastructure and the working-capital position of micro-entrepreneurs in Rivers State. Three research questions guided the study, using a descriptive survey design. Data were collected from 320 micro-entrepreneurs operating in Port Harcourt's major markets and its POS agent networks, selected through stratified random sampling, using a questionnaire rated on a four-point scale (Very High Extent, High Extent, Low Extent, Very Low Extent). Respondents reported a perceived shortening of the time between a sale and usable cash, a reduced amount of capital they feel they need to hold idle, and greater ease restocking inventory quickly, though the size of these reported effects depends heavily on how reliable an entrepreneur's settlement and liquidity access actually is. The weakest area, by some margin, is the perceived association between faster payments and access to formal credit: real-time transaction records are not yet reported as translating into better loan terms for the entrepreneurs generating them. Given the descriptive, cross-sectional survey design, these findings should be read as respondents' perceptions and reported associations rather than as evidence of causal impact. The study recommends that payment providers and regulators treat transaction-data portability as a deliberate policy goal, not an accidental byproduct of instant settlement.

TacticFix Artifact This artifact accompanies our study of tactic-aware vulnerability remediation. It contains the TacticFix implementation, benchmark, executable oracles, retrieval data, evaluation sc…

TacticFix Artifact This artifact accompanies our study of tactic-aware vulnerability remediation. It contains the TacticFix implementation, benchmark, executable oracles, retrieval data, evaluation scripts, prompts, and archived per-case results. The artifact includes: A canonical benchmark of 118 real-world vulnerability-remediation cases: 93 npm, 15 PyPI, and 10 Maven cases. Six remediation tactics: version adjustment, alternative libraries, bypassing, configuration changes, patching, and re-initialization. Executable vulnerability oracles and regression checks. Frozen GitHub issue context and a prebuilt ChromaDB retrieval index. Evaluation harnesses for TacticFix, general-purpose agents, LLM baselines, version-based tools, model backends, and ablation variants. Per-case outputs and documentation of benchmark, prompt, RAG, and result provenance. To validate the packaged benchmark, result files, oracle paths, and retrieval collections, run: python3 scripts/validate_artifact.py For the complete offline smoke test, including 455 oracle fault-injection checks, run: ./run_smoke_test.sh Fresh end-to-end evaluations require the corresponding language toolchains, model credentials, and cloned target repositories. Generated environments, third-party repository clones, package caches, and credentials are intentionally excluded. Please see README.md for setup and execution instructions, and docs/ for detailed benchmark, prompt, RAG, CodeQL, and result documentation.

This repository contains supplementary materials associated with the study on evaluation criteria for Extended Reality (XR) applications in industrial contexts. The repository includes:(1) the questio…

This repository contains supplementary materials associated with the study on evaluation criteria for Extended Reality (XR) applications in industrial contexts. The repository includes:(1) the questionnaire instrument used for the industrial and scientific panels; and(2) the anonymized questionnaire responses collected from both participant groups. These materials are provided to support transparency and reproducibility of the questionnaire-based analysis reported in the associated manuscript.

This dataset contains multi-scale vascular plant community data collected across a woody encroachment gradient in sub-Mediterranean grassland–shrubland mosaics in the northern Iberian Peninsula.…

This dataset contains multi-scale vascular plant community data collected across a woody encroachment gradient in sub-Mediterranean grassland–shrubland mosaics in the northern Iberian Peninsula. Vegetation data were recorded in 144 plots distributed across four study sites and six nested spatial scales: 0.0001, 0.001, 0.01, 0.1, 1 and 10 m². The dataset includes presence–absence matrices of vascular plant species for each spatial scale and plot, functional trait data for the selected plant traits, Ecological Indicator Values for Europe (EIVE) for the recorded species, and plot-level woody cover values. It also includes final summary tables with community-level metrics calculated for each plot and spatial scale, including community-weighted means of functional traits, community mean ecological indicator values, and woody cover information. These data were used to assess how spatial scale and woody cover jointly affect species richness, functional diversity, phylogenetic diversity, phylogenetic–functional decoupling and community ecological affinities along a grassland–shrubland gradient.

This study examines which institutional and market factors are associated with budget savings in public procurement across the regions of Uzbekistan. A balanced regional framework was constructed for …

This study examines which institutional and market factors are associated with budget savings in public procurement across the regions of Uzbekistan. A balanced regional framework was constructed for 14 territorial units over 2019–2024, yielding 84 region-year observations. The dependent variable is the budget-saving rate, measured as the percentage difference between planned and contracted procurement expenditure. Eight explanatory variables capture electronic procurement, participation by small and medium-sized enterprises, publication openness, bidder intensity, single-bid contracting, procurement-cycle duration, staff qualification, and the use of framework agreements. A regional fixedeffects model is estimated to isolate within-country variation while controlling for time-invariant regional characteristics.

Мақолада прокуратура органларининг коррупцияни олдини олиш ва унга қарши курашишда қонун ижодкорлиги фаолиятидаги иштироки, унинг асосий шакллари ҳамда ҳуқуқий асослари таҳлил қилинган. Прокуратура ор…

Мақолада прокуратура органларининг коррупцияни олдини олиш ва унга қарши курашишда қонун ижодкорлиги фаолиятидаги иштироки, унинг асосий шакллари ҳамда ҳуқуқий асослари таҳлил қилинган. Прокуратура органларининг қонун лойиҳаларини ишлаб чиқиш, муҳокама қилиш, ҳуқуқий ва коррупцияга қарши экспертизадан ўтказиш, шунингдек, қонунчиликни такомиллаштириш бўйича таклифлар киритишдаги иштироки коррупциявий хавфларни барвақт аниқлаш ва бартараф этишнинг превентив механизми сифатида ёритилган. Тадқиқот натижасида прокуратура органларининг қонун ижодкорлигидаги иштирокини янада такомиллаштириш, коррупцияга қарши экспертиза соҳасидаги ваколатларини ҳуқуқий жиҳатдан аниқлаштириш бўйича таклифлар илгари сурилган.

## Overview Overview Dataset obsahuje individuální odpovědi z průřezového online dotazníkového šetření u dospělých v České republice. Studie byla zaměřena na vývoj a psychometrické ověření čtyřpoložko…

## Overview Overview Dataset obsahuje individuální odpovědi z průřezového online dotazníkového šetření u dospělých v České republice. Studie byla zaměřena na vývoj a psychometrické ověření čtyřpoložkové škály Compulsive Pornography Use Scale (CPUS-4), která zachycuje behaviorální projevy kompulzivního užívání pornografie v návaznosti na rámec ICD-11 pro kompulzivní sexuální chování. Součástí dat jsou sociodemografické údaje, charakteristiky užívání pornografie a položky a skóry CPUS-4, PAST, DASS-21 a UCLA-3. Rukopis uvádí výchozí vzorek 1 134 respondentů a analytický vzorek 1 023 respondentů po vyloučení 111 případů s chybějícími položkami hlavních psychometrických nástrojů. Data slouží především k psychometrickým analýzám a studiu souvislostí mezi užíváním pornografie, psychickými obtížemi a osamělostí. Nereprezentativní výběr, sebevýpovědní měření a průřezový design omezují zobecnitelnost a neumožňují vyvozovat kauzalitu. CPUS-4 není samostatným diagnostickým nástrojem. Jedná se o citlivé osobní údaje, které mohou být zpřístupněny pouze na základě žádosti adresované autorům a s jejich souhlasem. Související článek: Babilonová, T., Pontes, H. M., Pipová, H., Reháková, T., Komrska, Š., & Suchá, J. (2026). The development and psychometric evaluation of the brief Compulsive Pornography Use Scale (CPUS-4). BMC Psychology. https://doi.org/10.1186/s40359-026-05328-1. ## Methods • Průřezové, neintervenční online dotazníkové šetření u dospělých v České republice. • Nepravděpodobnostní příležitostný výběr a samovýběr; distribuce přes sociální sítě Facebook, Instagram a Discord. • Sběr probíhal od ledna do března 2024. • Analytický vzorek: N = 1 023, 364 mužů, 632 žen, 10 osob v kategorii "jiné" a 17 chybějících údajů o genderu. Respondenti jsou ve věku 18–72 let, průměrný věk je 27,52 roku (SD = 7,74). • Pro EFA a CFA byl analytický vzorek náhodně rozdělen se stratifikací podle genderu a věkových tercilů: EFA n = 511, CFA n = 512. Chybějící gender tvořil samostatné stratum. ## Data structure Proměnná / proměnné Popis, jednotka a kódování ID Číselný identifikátor záznamu; 1 089 dostupných jedinečných hodnot. Date Datum a čas záznamu. Time Číselný údaj 146–9 249; doba vyplňování v sekundách. Gender 1 = muž, 2 = žena, 4 = jiné; mapování odvozené dle výše uvedeného postupu. Kód 3 (nechci odpovědět) se nevyskytuje. Age Věk v letech; 18–72. occupation 0 = neuvedeno; 1 = student/ka; 2 = zaměstnaný/á; 3 = OSVČ; 4 = nezaměstnaný/á / nestudující; 5 = jiné (např. rodičovská dovolená, důchod, dlouhodobá péče). education 0 = neuvedeno; 1 = základní (ISCED 2); 2 = střední bez maturity (ISCED 3); 3 = střední s maturitou (ISCED 3); 4 = vyšší odborné (ISCED 5); 5 = vysokoškolské (ISCED 6–8). ageporn Odhadovaný věk prvního kontaktu s pornografickým materiálem, roky; 0–40. Nula znamená nikdy. porn30d Počet případů sledování pornografie za posledních 30 dnů; 0–50. Jde o počet sledování, nikoli počet dnů. porntime1 Průměrná denní doba sledování pornografie v pracovní den, v hodinách (rozpětí 0–5) porntime2 Průměrná denní doba sledování pornografie ve volný den, v hodinách (rozpětí 0–12). ucla1, ucla2, ucla3 Položky osamělosti; v souboru 0, 1, 2, 3. Nula znamená, že respondent položku nevyplnil. Odpovědi byly kódovány na škále od 1 do 3 následovně: 1 = téměř nikdy, 2 = někdy a 3 = často. ucla_ts Součet položek ucla1–ucla3. past1–past25 Každý celočíselný index od 1 do 25 označuje samostatnou položku PAST. Odpovědi byly kódovány na škále od 0 do 4 následovně: 0 = nikdy, 1 = jeden až dvakrát, 2 = zřídka, 3 = někdy a 4 = často. past_ts Součet položek past1–past25. cpus1, cpus2, cpus3, cpus4 Položky CPUS-4; Odpovědi byly kódovány na škále od 1 do 5 následovně: 1 = nikdy, 2 = málokdy, 3 = někdy, 4 = často a 5 = velmi často. Nula znamená, že respondent položku nevyplnil. cpus_ts Součet položek cpus1–cpus4 dass1, dass6, dass8, dass11, dass12, dass14, dass18 Sedm položek faktoru stresu DASS-21. Odpovědi byly kódovány na škále od 0 do 3 následovně: 0 = vůbec se mě to netýkalo, 1 = do určité míry nebo někdy se mě to týkalo, 2 = do velké míry nebo často se mě to týkalo a 3 = velice nebo skoro vždy se mě to týkalo. dass3, dass5, dass10, dass13, dass16, dass17, dass21 Sedm položek faktoru deprese DASS-21. Odpovědi byly kódovány na škále od 0 do 3 následovně: 0 = vůbec se mě to netýkalo, 1 = do určité míry nebo někdy se mě to týkalo, 2 = do velké míry nebo často se mě to týkalo a 3 = velice nebo skoro vždy se mě to týkalo. dass2, dass4, dass7, dass9, dass15, dass19, dass20 Sedm položek faktoru úzkosti DASS-21. Odpovědi byly kódovány na škále od 0 do 3 následovně: 0 = vůbec se mě to netýkalo, 1 = do určité míry nebo někdy se mě to týkalo, 2 = do velké míry nebo často se mě to týkalo a 3 = velice nebo skoro vždy se mě to týkalo. dep_ts Součet sedmi položek faktoru deprese. anx_ts Součet sedmi položek faktoru úzkosti. str_ts Součet sedmi položek faktoru stresu. ## Usage Jak data načíst Příklad pro Python s pandas a openpyxl. Soubory umístěte do pracovního adresáře. Kód zachovává původní hodnoty a neprovádí neověřené překódování. Příklad pro Python: from pathlib import Path import pandas as pd path = Path("dataset_cpus.xlsx") df = pd.read_excel(path, sheet_name="Data", engine="openpyxl") df = df.dropna(axis=1, how="all") # pouze zcela prázdné sloupce print("Řádkové pozice a proměnné:", df.shape) # (1134, 72) print("Zcela prázdné řádky:", df.isna().all(axis=1).sum()) # 14 items = ( [f"ucla{i}" for i in range(1, 4)] + [f"past{i}" for i in range(1, 26)] + [f"cpus{i}" for i in range(1, 5)] + [f"dass{i}" for i in range(1, 22)] ) analytic = df.dropna(subset=items).copy() print("Kompletní případy:", len(analytic)) # 1023 # Jde o reprodukci výběru podle chybějících položek, # nikoli potvrzení správnosti veškerého kódování. print("Nuly v UCLA v analytickém vzorku:", analytic[["ucla1", "ucla2", "ucla3"]].eq(0).any(axis=1).sum()) # 6 # Přepočet do nového sloupce, bez přepsání uloženého skóru. past_items = [f"past{i}" for i in range(1, 26)] df["past_ts_recomputed"] = df[past_items].sum(axis=1, min_count=25) available = df[["past_ts", "past_ts_recomputed"]].notna().all(axis=1) mismatch = available & df["past_ts"].ne(df["past_ts_recomputed"]) print(df.loc[mismatch, ["ID", "past_ts", "past_ts_recomputed"]]) Požadavky na software nebo závislosti Pro uvedený příklad je potřeba Python 3, pandas a openpyxl (pip install pandas openpyxl). Pro reprodukci publikovaných analýz je relevantní R a balíčky uvedené v oddílu Methods. XLSX lze prohlížet také v Excelu nebo LibreOffice Calc. Známá omezení pro re-use • Prázdné buňky znamenají, že údaj byl odstraněn v rámci čištění dat. • Výběr kompletních případů reprodukuje velikost analytického vzorku a tabulku 1. • Vzorek je nereprezentativní a převažují ženy a mladší dospělí. Nelze z něj bez dalšího odvozovat populační prevalenci ani zobecňovat výsledky na děti, klinickou populaci nebo jiné jazykové prostředí. • Sebevýpovědní údaje mohou být ovlivněny vybavováním a sociální žádoucností. Jednotlivé nástroje mají rozdílná referenční období. • Použití datasetu je podmíněno vyžádáním od korespondenčního autora, jeho licence je CC BY 4.0. • Před veřejným sdílením je nutné vyjasnit zacházení s přesnými časovými značkami a dalšími potenciálně rozlišujícími údaji. ## Change log - 2026-09-26: První vydání ## Contact Jaroslava Suchá, jaroslava.sucha@upol.cz ACCESS CONDITIONS Access to this record is restricted due to: Personal sensitive data. Access may be granted for legitimate research purposes upon reasonable request and subject to approval by the data steward. To request access, please contact: Jaroslava Suchá (jaroslava.sucha@upol.cz). The request should include: • brief project description and intended use • institutional affiliation • agreement to sign a Data Use Agreement (DUA), if required