To adapt to the intensive, bursty, and latency-constrained traffic from large language model (LLM) training and inference, scale-up networks are now facing tremendous challenges. Existing electrical packet switching (EPS) fabrics provide packet-level flexibility at the cost of high power consumption and long latency. Introducing optical circuit switching (OCS) in scale-up networks offers direct optical connections that can effectively reduce power consumption and latency, but OCS lacks the packet-level flexibility required by LLM inference (especially for mixture-of-experts (MoE) inference). To address these dilemmas, this work presents xSwitch, an optical-electrical-integrated (O-E-integrated) interconnect for scale-up networks, and demonstrates its effectiveness experimentally. Unlike the traditional hybrid-optical-electrical interconnects that usually place EPS and OCS in parallel, xSwitch integrates one port-count-reduced (defined by the O/E port ratio) EPS layer (ESL) on top of an OCS layer (OSL). Then, OSL can establish optical connections for xPU pairs with stable and intensive traffic demands, while bypassing the ESL for energy and latency reduction, and the dynamic and unpredictable traffic between xPUs can be provisioned by letting OSL forward it to the ESL. We prototype xSwitch with off-the-shelf components, validate its effectiveness with real-world MoE inference tasks, and also confirm its scalability with large-scale simulations. Our results indicate that for MoE inference workloads, xSwitch with a 2:1 O/E port ratio limits the average gaps to the full-EPS baseline to 1.96% in TTFT and 0.59% in TPOT.

This paper introduces Graphon Attachment Network Models (GAN-M), a novel framework for modeling evolving networks with rich structural dependencies, grounded in graphon theory. GAN-M provides a flexible and interpretable foundation for studying network formation by leveraging graphon functions to define attachment probabilities, thereby combining the strengths of graphons with a temporal perspective. A key contribution of this work is a methodology for learning structural changes in these networks over time. Our approach uses graph counts—frequencies of substructures such as triangles and stars—to capture shifts in network topology. We propose a new statistic designed to learn changes in the resulting piecewise polynomial signals and develop an efficient method for change detection, supported by theoretical guarantees. Numerical experiments demonstrate the effectiveness of our approach across various network settings, highlighting its potential for dynamic network analysis.

Traditional network inference methods, such as Gaussian Graphical Models, which are built on continuity and homogeneity, face challenges when modeling discrete data and heterogeneous frameworks. Furthermore, under high-dimensionality, the parameter estimation of such models can be hindered by the notorious intractability of high-dimensional integrals. In this paper, we introduce a new and flexible device for graphical models, which accommodates diverse data types, including Gaussian, Poisson log-normal, and latent Gaussian copula models. The new device is driven by a new marginally recoverable parametric family, which can be effectively estimated without evaluating the high-dimensional integration in high-dimensional settings thanks to the marginal recoverability. We further introduce a mixture of marginally recoverable models to capture ubiquitous heterogeneous structures. We show the validity of the desirable properties of the models and the effective estimation methods, and demonstrate their advantages over the state-of-the-art network inference methods via extensive simulation studies and a gene regulatory network analysis of real single-cell RNA sequencing data.

Graphons offer a powerful framework for modeling large-scale networks, yet estimation remains challenging. We propose a novel approach that leverages a low-rank additive representation, yielding both a low-rank connection probability matrix and a low-rank graphon--two goals rarely achieved jointly. Our method resolves identification issues and enables an efficient sequential algorithm based on subgraph counts and interpolation. We establish consistency and demonstrate strong empirical performance in terms of computational efficiency and estimation accuracy through simulations and data analysis.

This work presents the first theoretically justified simultaneous inference framework for off-policy evaluation (OPE). In contrast to existing methods that focus on point estimates or pointwise confidence intervals (CIs), the new framework quantifies global uncertainty across an infinite or continuous initial state space, offering valid inference over the entire state space. Our method leverages sieve-based Q-function estimation and (high-dimensional) Gaussian approximation techniques over convex regions, which further motivates a new multiplier bootstrap algorithm for constructing asymptotically correct simultaneous confidence regions (SCRs). The widths of the SCRs exceed those of the pointwise CIs by only a logarithmic factor, indicating that our procedure is nearly optimal in terms of efficiency. The effectiveness of the proposed approach is demonstrated through simulations and analysis of the OhioT1DM dataset.

This paper proposes a unified framework for inferring large-scale time-varying correlation networks via data-driven time-varying thresholds that can control uncertainty simultaneously. The framework allows the dimension of time series vectors to be fixed or diverging at a high polynomial rate of the sample size. It also allows the time series to exhibit changing temporal characteristics beyond stationarity without specific structural assumptions. Motivated by the practical issue that the confidence band of non-parametric estimators of correlations can exceed their natural domain [−1, 1] (see, for example [1])), we propose a simple uniform variance reduction technique. When applied to the construction of a correlation network, the new device yields more accurate thresholds, which enhance the probability of recovering the time-varying network structures. We broaden the applicability of our method by developing difference-based estimators of cross-correlations that are robust to structure breaks in the time-varying mean functions, and by allowing both a fixed and a diverging number of lags in the correlation functions. We prove the asymptotic validity of the proposed method, especially in achieving accurate family-wise error control when disclosing flexible time-varying network structures. The effectiveness of our method in finite samples is demonstrated through simulation studies and data analysis.