Dmitry Zaytsev

Time-varying causal models provide a powerful framework for studying dynamic scientific systems, yet most existing approaches assume that the underlying causal network is known a priori - an assumption rarely satisfied in real-world domains where causal structure is uncertain, evolving, or only indirectly observable. This limits the applicability of dynamic causal inference in many scientific settings. We propose Dynamic Causal Network Autoregression (DCNAR), a two-stage neural causal modeling framework that integrates data-driven causal discovery with time-varying causal inference. In the first stage, a neural autoregressive causal discovery model learns a sparse directed causal network from multivariate time series. In the second stage, this learned structure is used as a structural prior for a time-varying neural network autoregression, enabling dynamic estimation of causal influence without requiring pre-specified network structure. We evaluate the scientific validity of DCNAR using behavioral diagnostics that assess causal necessity, temporal stability, and sensitivity to structural change, rather than predictive accuracy alone. Experiments on multi-country panel time-series data demonstrate that learned causal networks yield more stable and behaviorally meaningful dynamic causal inferences than coefficient-based or structure-free alternatives, even when forecasting performance is comparable. These results position DCNAR as a general framework for using AI as a scientific instrument for dynamic causal reasoning under structural uncertainty.

This paper advances a methodological framework for the study of policy polarization. We conceptualize polarization as the structure of relationships among political and epistemic actors, measurable through the similarity of their policy statements. To generate these relational data, large language models (LLMs) are employed as scalable tools for extracting, summarizing, and synthesizing policy statements or positions from diverse textual sources, including debates, party platforms, interviews, and think tank publications. The resulting corpus is embedded into a shared vector space, and similarity measures between statements provide the basis for constructing multimodal networks that connect policy statements through semantic similarity and link them to candidates and think tanks. These networks enable the application of social network analysis, such as correspondence analysis, influence modeling, and structural comparison, to model alignments, divergences, and patterns of polarization. The framework contributes a replicable and extensible approach to analyzing policy polarization, integrating computational text processing with network-analytic models to capture the relational dynamics of political discourse.

Nonlinear machine-learning models are increasingly used to discover causal relationships in time-series data, yet the interpretation of their outputs remains poorly understood. In particular, causal scores produced by regularized neural autoregressive models are often treated as analogues of regression coefficients, leading to misleading claims of statistical significance. In this paper, we argue that causal relevance in nonlinear time-series models should be evaluated through forecast necessity rather than coefficient magnitude, and we present a practical evaluation procedure for doing so. We present an interpretable evaluation framework based on systematic edge ablation and forecast comparison, which tests whether a candidate causal relationship is required for accurate prediction. Using Neural Additive Vector Autoregression as a case study model, we apply this framework to a real-world case study of democratic development, modeled as a multivariate time series of panel data - democracy indicators across 139 countries. We show that relationships with similar causal scores can differ dramatically in their predictive necessity due to redundancy, temporal persistence, and regime-specific effects. Our results demonstrate how forecast-necessity testing supports more reliable causal reasoning in applied AI systems and provides practical guidance for interpreting nonlinear time-series models in high-stakes domains.

This paper examines Web3 ecosystems not merely as markets for digital assets, but as networked social spaces where economic transactions give rise to enduring social ties, shared narratives, and collective identities. Leveraging large-scale data mining of fused on-chain blockchain transactions and off-chain social media activity, we analyze over one hundred NFT collections to uncover how different forms of participation structure community formation in decentralized environments. Using network analysis, we identify distinct ecosystem roles, such as long-term holders, active traders, and short-term speculators, and demonstrate how each produces markedly different network topologies, levels of cohesion, and pathways for influence. We complement this structural analysis with discourse analysis of social media engagement, revealing how narrative production, visibility, and sustained interaction persist even as transactional activity declines. Our findings show that communities centered on holding behavior evolve from transactional networks into socially embedded ecosystems characterized by dense ties, decentralized influence, and ongoing cultural participation, while trader- and speculator-dominated networks remain fragmented and transactional. By linking network structure with discursive dynamics, this study provides a sociotechnical framework for understanding how value, identity, and inequality are negotiated in Web3 spaces. The approach offers a scalable method for detecting patterns of inclusion, exclusion, and representational imbalance, advancing network-based research on digital communities beyond purely economic or technical accounts.

The intersection of Artificial Intelligence (AI) and democracy has emerged as a significant yet underexplored domain in academic research, despite its prominence in public discourse. This study investigates the state of the art in “AI and Democracy” by analyzing 745 academic articles from the Web of Science (WoS) and 7,058 media records from Factiva over a period spanning more than 40 years. Using advanced text mining and topic modeling techniques, we identify key themes in academic and public discourse, highlighting overlaps and gaps between the two. Our findings reveal that while academia focuses on AI applications for public policy, democratic development, and complex modeling of democratic phenomena, public discourse emphasizes governance, regulation, and the social impact of AI, often reflecting concerns about misinformation, disinformation, and threats. This research underscores the need for greater alignment between academic inquiry and public priorities to address pressing social challenges, foster interdisciplinary collaboration, and ensure that AI serves as a force for democratic resilience and development.

In this paper, we answer the multiple calls for systematic analysis of paradigms and subdisciplines in political science—the search for coherence within a fragmented field. We collected a large dataset of over seven hundred thousand writings in political science from Web of Science since 1946. We found at least two waves of political science development, from behaviorism to new institutionalism. Political science appeared to be more fragmented than literature suggests—instead of ten subdisciplines, we found 66 islands. However, despite fragmentation, there is also a tendency for integration in contemporary political science, as revealed by co-existence of several paradigms and coherent and interconnected topics of the “canon of political science,” as revealed by the core-periphery structure of topic networks. This was the first large-scale investigation of the entire political science field, possibly due to newly developed methods of bibliometric network analysis: temporal bibliometric analysis and island methods of clustering. Methodological contribution of this work to network science is evaluation of islands method of network clustering against a hierarchical cluster analysis for its ability to remove misleading information, allowing for a more meaningful clustering of large weighted networks.

In this paper, we examine the contribution of Network Science journal to the network science discipline. We do so from two perspectives. First, expanding the existing taxonomy of article contribution, we examine trends in theory testing, theory building, and new method development within the journal’s articles. We find that the journal demands a high level of theoretical contribution and methodological rigor. High levels of theoretical and methodological contribution become significant predictors of article citation rates. Second, we look at the composition of the studies in Network Science and determine that the journal has already established a solid “hard core” for the new discipline.