Selected work

Representative Papers

Comorbidity Network Analysis Reveals Diagnostic Disparities Between Austrian and Non-Austrian Inpatients: A Population-Wide Cohort Study

Jul 05, 2026

This study investigates diagnostic inequities in multimorbidity among international migrants attributable to barriers in healthcare access. Leveraging approximately 13 million hospital records from Austria between 2015 and 2019, the authors constructed a 1:1 matched cohort of Austrian-born and foreign-born patients using propensity score matching. Through multimorbidity network modeling and statistical comparison, they reveal for the first time that migrant patients exhibit significantly fewer multimorbidity connections—particularly between mental and metabolic conditions—not due to lower actual disease burden, but likely reflecting structural healthcare barriers. The analysis further uncovers gender-specific patterns: migrant women show clustering of depression, somatization, and back pain, whereas migrant men are characterized by a higher incidence of acute somatic conditions, highlighting systemic disparities in diagnostic recognition.

0 citationsRead paper

Who Brought Easter Eggs to Eid? Auditing Cultural Translation of Math Word Problems Across Diverse Languages and Regions

Jun 09, 2026

This study addresses persistent challenges in cross-lingual translation of mathematical word problems by large language models, particularly insufficient cultural consistency, diversity compression, and misjudgment of regional context. Through the first large-scale corpus analysis, the authors systematically audit how Claude Opus 4, GPT-4.1, and Gemini 2.5 Pro handle culturally embedded entities—such as names, foods, and locations—when translating 60 English problems into seven languages spanning high- and low-resource settings. Combining human annotation with quantitative evaluation, they perform fine-grained coding of 6,489 cultural transformation instances, revealing widespread entropy collapse in diversity, surface-level token preferences, systematic regional misattribution, and frequent cross-cultural contamination errors (e.g., “Easter eggs used in Eid celebrations”). The work introduces the first fine-grained annotation framework for cultural translation, finding model agreement on transformation type in only 62.5% of cases and exact substitution alignment in just 33.5%.

0 citationsRead paper

Reconstructing temporal multi-relational firm networks at scale using large language models. The case of the semiconductor industry

May 15, 2026

This study addresses the challenge of capturing the dynamic complexity of global multi-relational corporate networks in the semiconductor industry, which traditional proprietary databases—often costly and lagging—struggle to track in a timely manner. The authors propose a generalizable framework that leverages large language models to automatically extract and classify supply chain, collaboration, and ownership relationships from 170 million open web pages, constructing a time-series multi-relational network encompassing over 1,300 firms. Integrating web crawling, natural language processing, and graph analytics, the approach enables high-tempo, automated structuring of relational data, achieving a precision of 0.884 and an F1 score of 0.784 in link extraction. Empirical analysis reveals network contraction during the 2022 chip shortage, a sharp rise in centrality among AI-critical firms, and geographically reconfigured ties driven by geopolitical forces.

0 citationsRead paper

Emergent Self-Attention from Astrocyte-Gated Associative Memory Dynamics

Apr 28, 2026

This work addresses the limitations of traditional associative memory models, which suffer from degraded retrieval performance under high memory load and interference and lack a dynamical systems explanation for self-attention mechanisms. The authors propose a novel Hopfield-type associative memory model incorporating astrocyte-regulated neuronal gain, whose dynamics are governed by an entropy-regularized replicator equation. This formulation naturally yields softmax-normalized pattern similarity allocation over the gain simplex. The resulting coupled system exhibits global convergence and, for the first time, reveals self-attention as an emergent routing behavior modulated by astrocytes from a dynamical systems perspective. Experimental results demonstrate that the proposed model significantly outperforms classical Hopfield networks and existing neuro-glial baselines under conditions of high memory load and strong interference, achieving markedly higher retrieval accuracy.

0 citationsRead paper

Car Dependency in Urban Accessibility

Apr 01, 2026

This study addresses the dual challenges of rising transport-related carbon emissions and mobility inequity driven by excessive reliance on private vehicles in urban areas. It proposes the first quantifiable and scalable Car Dependency Index (CDI), integrating high-resolution geospatial data with accessibility modeling to analyze 18 cities across Europe and North America. Using counterfactual scenario simulations, the research evaluates the impact of public transit network expansions, revealing that car dependency remains a key determinant of vehicle ownership even after controlling for income. Systemic upgrades—such as Rome’s metro expansion—can substantially reduce commuting cars (by approximately 60,000), whereas isolated interventions yield limited benefits. The proposed framework enables precise identification of car-free zones and priority areas for transit investment, offering a robust foundation for equitable, low-carbon urban transport planning.

0 citationsRead paper

Recent publications

Latest Papers

Comorbidity Network Analysis Reveals Diagnostic Disparities Between Austrian and Non-Austrian Inpatients: A Population-Wide Cohort Study

Jul 05, 2026

This study investigates diagnostic inequities in multimorbidity among international migrants attributable to barriers in healthcare access. Leveraging approximately 13 million hospital records from Austria between 2015 and 2019, the authors constructed a 1:1 matched cohort of Austrian-born and foreign-born patients using propensity score matching. Through multimorbidity network modeling and statistical comparison, they reveal for the first time that migrant patients exhibit significantly fewer multimorbidity connections—particularly between mental and metabolic conditions—not due to lower actual disease burden, but likely reflecting structural healthcare barriers. The analysis further uncovers gender-specific patterns: migrant women show clustering of depression, somatization, and back pain, whereas migrant men are characterized by a higher incidence of acute somatic conditions, highlighting systemic disparities in diagnostic recognition.

0 citationsRead paper

Who Brought Easter Eggs to Eid? Auditing Cultural Translation of Math Word Problems Across Diverse Languages and Regions

Jun 09, 2026

This study addresses persistent challenges in cross-lingual translation of mathematical word problems by large language models, particularly insufficient cultural consistency, diversity compression, and misjudgment of regional context. Through the first large-scale corpus analysis, the authors systematically audit how Claude Opus 4, GPT-4.1, and Gemini 2.5 Pro handle culturally embedded entities—such as names, foods, and locations—when translating 60 English problems into seven languages spanning high- and low-resource settings. Combining human annotation with quantitative evaluation, they perform fine-grained coding of 6,489 cultural transformation instances, revealing widespread entropy collapse in diversity, surface-level token preferences, systematic regional misattribution, and frequent cross-cultural contamination errors (e.g., “Easter eggs used in Eid celebrations”). The work introduces the first fine-grained annotation framework for cultural translation, finding model agreement on transformation type in only 62.5% of cases and exact substitution alignment in just 33.5%.

0 citationsRead paper

Reconstructing temporal multi-relational firm networks at scale using large language models. The case of the semiconductor industry

May 15, 2026

This study addresses the challenge of capturing the dynamic complexity of global multi-relational corporate networks in the semiconductor industry, which traditional proprietary databases—often costly and lagging—struggle to track in a timely manner. The authors propose a generalizable framework that leverages large language models to automatically extract and classify supply chain, collaboration, and ownership relationships from 170 million open web pages, constructing a time-series multi-relational network encompassing over 1,300 firms. Integrating web crawling, natural language processing, and graph analytics, the approach enables high-tempo, automated structuring of relational data, achieving a precision of 0.884 and an F1 score of 0.784 in link extraction. Empirical analysis reveals network contraction during the 2022 chip shortage, a sharp rise in centrality among AI-critical firms, and geographically reconfigured ties driven by geopolitical forces.

0 citationsRead paper

Emergent Self-Attention from Astrocyte-Gated Associative Memory Dynamics

Apr 28, 2026

This work addresses the limitations of traditional associative memory models, which suffer from degraded retrieval performance under high memory load and interference and lack a dynamical systems explanation for self-attention mechanisms. The authors propose a novel Hopfield-type associative memory model incorporating astrocyte-regulated neuronal gain, whose dynamics are governed by an entropy-regularized replicator equation. This formulation naturally yields softmax-normalized pattern similarity allocation over the gain simplex. The resulting coupled system exhibits global convergence and, for the first time, reveals self-attention as an emergent routing behavior modulated by astrocytes from a dynamical systems perspective. Experimental results demonstrate that the proposed model significantly outperforms classical Hopfield networks and existing neuro-glial baselines under conditions of high memory load and strong interference, achieving markedly higher retrieval accuracy.

0 citationsRead paper

Car Dependency in Urban Accessibility

Apr 01, 2026

This study addresses the dual challenges of rising transport-related carbon emissions and mobility inequity driven by excessive reliance on private vehicles in urban areas. It proposes the first quantifiable and scalable Car Dependency Index (CDI), integrating high-resolution geospatial data with accessibility modeling to analyze 18 cities across Europe and North America. Using counterfactual scenario simulations, the research evaluates the impact of public transit network expansions, revealing that car dependency remains a key determinant of vehicle ownership even after controlling for income. Systemic upgrades—such as Rome’s metro expansion—can substantially reduce commuting cars (by approximately 60,000), whereas isolated interventions yield limited benefits. The proposed framework enables precise identification of car-free zones and priority areas for transit investment, offering a robust foundation for equitable, low-carbon urban transport planning.

0 citationsRead paper