FRB 20220912A is a highly active repeating fast radio burst (FRB) source, discovered by the Canadian Hydrogen Intensity Mapping Experiment (CHIME) using its real-time FRB detection system (CHIME/FRB). Here, we present results from a radio monitoring campaign of FRB 20220912A using CHIME, including ~200 hours of data collected by CHIME/Pulsar, spanning 1.5 years following the source's discovery. We present an analysis of a sample of 828 CHIME-detected bursts from FRB 20220912A, in the 400-800 MHz radio frequency band. The source remains highly active for ~10 weeks and has a bimodal wait-time distribution with peaks at $160^{+120}_{-70}$ ms and $306^{+14}_{-13}$ s. Assuming a radio efficiency factor of $10^{-4}$ and a beaming angle of 0.1, we estimate the total emitted energy from the source over the entire observing campaign to be $2 \times 10^{43}$ ergs. We report a 2.3$σ$ detection of a linear increase in the DM of $1.4 \pm 0.6$ pc cm$^{-3}$ yr$^{-1}$, with no significant trend in rotation measure (with a 3$σ$ upper limit of 13.4 rad m$^{-2}$ yr$^{-1}$). We contrast our findings with other active repeaters, which exhibit different DM and RM evolution to indicate that FRB 20220912A may reside in a unique local environment.

We propose ultrafast all-optical switching exploiting the bistability between a spatially uniform photon superfluid and a spontaneously ordered supersolid in a driven-dissipative microcavity. The key ingredient is a tunable nonlocal photon--photon interaction engineered by embedding a high-mobility two-dimensional electron gas (2DEG) inside the cavity. A drift current displaces the Fermi disk, imparting a negative region to the Lindhard interaction kernel at finite wavevectors and triggering a roton instability. The resulting bistable $S$-curve supports a write--hold--erase protocol in which short optical pulses toggle the system between branches with a switching contrast of order 120~dB. The hysteretic ON state persists under a constant sub-threshold drive after the write pulse is removed, realizing an all-optical bistable memory. Since the photon field couples additively to each embedded quantum well, stacking layers with distinct drift angles allows the roton profile to be engineered with higher-order symmetries, imprinting richer spatial order on the supersolid and enabling nonbinary generalizations of the switch. Operating in the ultrafast, sub-fJ regime, this platform outperforms most existing all-optical switches in contrast and reconfigurability.

We consider the sequence of independent centered Gaussian random elements of a separable Banach space and their consecutive closed convex hulls. If inicial elements converge weakly to some limite, then, as shown in Davydov- Paulauskas (2024), its normalized convex hulls converge, with probability 1, to the concentration ellipsoid of the limiting distribution. The goal of the present note is to show that if the assumption of weak convergence of the initial sequence is relaxed, than the limit set can be an arbitrary convex compact set.

We report a cryogenic buffer-gas-cell-aerodynamic-lens-stack setup that enables the generation of shock-frozen, dense, and controllable beams of various nanoparticles in the gas phase, including small and low-density species such as isolated proteins. We demonstrate characterization of the setup using strong-field ionization combined with velocity-map imaging, allowing the unambiguous detection of nanoparticles in the protein-size range and full reconstruction of the particle beams including determination of particle flux and number density. The generation and characterization workflow presented here provides a valuable approach for protein-like sample preparation and delivery in single-particle diffractive imaging, microscopy, and low-temperature nanoscience.

We study the nature of chaos in finite and infinite dimensional systems through a comparison between the Kuramoto Sivashinsky (KS) equation, the Lorenz system, and a Lorenz type reduction of the KS equation proposed by Wilczak. Numerical simulations of the KS equation reveal intrinsic spatio temporal chaos, with disorder evolving simultaneously in space and time. In contrast, the Lorenz system and the Wilczak reduction exhibit low dimensional temporal chaos lacking spatial complexity. Lyapunov exponent analysis highlights the finite-dimensional convergence properties of the reduced systems and underscores the fundamentally different dynamical nature of chaos in the KS equation. In particular, we demonstrate that low-dimensional reductions may reproduce transient chaotic signatures but do not necessarily retain the structural properties of infinite-dimensional dissipative systems.

Zeolitic Imidazolate Frameworks (ZIFs) are a family of metal--organic frameworks that feature metal centers tetrahedrally linked to imidazole-based ligands and adopt zeolite-like topologies. ZIFs formed by Zinc cations and imidazolate linkers exhibit a remarkable degree of polymorphism, which can be modulated by varying synthesis parameters or thermodynamic conditions (i.e., temperature and pressure). Computer simulations provide a unique way of studying these materials and their phase transitions from the microscopic standpoint, revealing their underlying molecular mechanisms. However, studying these mechanisms requires to be able to classify the phase of each molecular entity in an agnostic and automatic fashion, which is particularly challenging when the two phases involved are structurally very similar. In this work, we systematically study neural network classifiers to classify ZIF phases on-the-fly during molecular dynamics simulations. We test a variety of input features, differing both in the dimensionality and nature of the descriptors and in the kind of force field used for building the training/testing database. We reveal that even with low-dimensional descriptors the classification is highly accurate, while the use of high-dimensional descriptors leads to an even better performance. Training the classifier with configurations coming from different force fields we can remove force field bias and enhance the classifier performance and general applicability. Finally, we apply our classifiers to reveal mechanistic details of the ZIF-4-cp $\xrightarrow{}$ ZIF-4-cp-II phase transition.

Deploying large language models (LLMs) on mobile devices is an emerging trend to enable data privacy and offline accessibility of LLM applications. Modern mobile neural processing units (NPUs) make such deployment increasingly feasible. However, existing mobile LLM inference frameworks suffer from high start-up latency due to their inevitable cold starts, i.e., launching LLM inferences when the model is not hosted in device memory. In this paper, we identify the key bottleneck of mobile LLM cold starts as the waste of flash bandwidth on unimportant model parameters. We design EdgeFlow, a mobile LLM inference framework that mitigates the cold start issue by adaptively adjusting the precisions of LLM parameters. Specifically, EdgeFlow leverages 1) an NPU-aware adaptive quantization algorithm that assigns different precisions to weights in a finer granularity according to their importance and NPU constraints, 2) an SIMD-friendly packing format that accelerates the transformation of various-precision weights into fixed-sized NPU-native data types, and 3) a synergistic granular pipeline that coordinates CPU and NPU computation in a fine-grained and dynamic manner. Experimental results show that EdgeFlow reduces cold-start latency by up to 4.07x compared with three state-of-the-art mobile LLM inference frameworks, i.e., llama.cpp, MNN, and llm.npu, under comparable model accuracy.

We extracted a list of 662 nearby (within $\sim16$ Mpc) HI-detection sources from the Five-hundred-meter Aperture Spherical radio Telescope (FAST) All Sky HI Survey (FASHI) and made a visual identification of them with optical counterpart. This inspection led to the discovery of 71 new dwarf galaxies. All of them are dwarf irregular galaxies with ongoing star formation. They are characterized by the following median parameters: visual magnitude of $g=17.8$ mag and color $(g-r)=0.29$ mag, HI-flux $S_\mathrm{HI}=718$ mJy km/s, HI-mass $M_\mathrm{HI}=3.7\times10^7$ $M_\odot$, as well as the HI line-width of $W_{50}=37$ km/s.

We present a systematic study of gravitational wave (GW) signals from phase transitions and topological defects in a unified high-quality axion framework. The gauged $U(1)_g$ symmetry forbids any bias term that could lift the vacuum degeneracy, restricting the theory to the phenomenologically viable case $N_{\rm DW}=1$. Requiring the axion to account for the observed dark matter (DM) abundance and satisfy the high-quality condition constrains the gauge symmetry-breaking scale to $f_g \in [1.6\times10^{11},\,10^{16}]\,\mathrm{GeV}$ for the QCD axion, leading to a well-defined band of GW signals, part of which is consistent with current pulsar timing array observations. Two-step first-order phase transitions are common in this framework, with the lower-scale transition generating GWs with $f^{\rm peak} \gtrsim \mathcal{O}(10^7)\,\mathrm{Hz}$. For axion-like realizations, generic post-inflation models predict GW spectra that are nearly degenerate with the QCD axion case. We conclude that GWs alone cannot distinguish between these scenarios, highlighting the need for complementary probes.

We continued the search for single pulses (SPs) in the northern part of the all-sky High Time Resolution Universe survey, whose aim is to detect pulsars and other radio transients. This search is now about 21% complete and has yielded the first discovery of a fast radio burst (FRB) with the 100 m Effelsberg Radio Telescope. FRB20110220A was detected with an S/N-optimised dispersion measure of 501.0 pc/cm$^{3}$ and a width of 11.9 $\pm$ 3.5 ms, for a fluence of 0.6 $\pm$ 0.1 Jy ms. We obtained the first L-band detection of the rotating radio transient (RRAT) J2028+28, from which we obtained upper limits on the source's period and burst rate, as well as an improved position. We also discovered a new RRAT, J0404+53, which had previously been reported as an isolated SP candidate. Eight new SP trains and 272 faint isolated SP candidates were detected too. We used these candidates to demonstrate that their all-sky detection rates depend on Galactic latitude and longitude. This direction dependence suggests the existence of a faint Galactic SP population.

We propose an adaptive control protocol for identifying the topology of dynamical networks interconnected over undirected graphs with cooperative and antagonistic interactions. The signed network is modeled using a repelling Laplacian. Topology identification relies on an edge-based formulation of the network and adaptive control protocols through the design of a persistently excited auxiliary network. Our approach guarantees the simultaneous identification and synchronization of the unknown signed network and establishes uniform semiglobal practical asymptotic stability of the estimation errors. Numerical simulations validate our theoretical results.

We study community detection in stochastic block models under pure node-level differential privacy, a stringent notion that protects the participation of an individual together with all of their incident edges. This setting is substantially more challenging than edge-private community detection, since modifying a single node can affect linearly many observations. On the algorithmic side, we analyze a node-private estimator based on the exponential mechanism combined with an extension lemma, and show that exact recovery remains achievable. In the standard sparse regime with logarithmic average degree and a fixed number of communities, our results imply that a logarithmic privacy budget suffices to obtain nontrivial recovery guarantees. On the lower bound side, we show that this logarithmic scaling is in fact unavoidable: any pure node-private method must fail to achieve polynomially small exact-recovery error, or polynomially small expected mismatch, unless the privacy budget is at least of this order. Moreover, in the regime of super-logarithmic privacy budgets, our upper and lower bounds yield a matching two-term characterization of the minimax risk, with one term governed by the non-private statistical signal and the other by the privacy budget; these match up to universal constants in the exponents. Taken together, our results identify an inherent logarithmic privacy cost in node-private community detection, absent under edge differential privacy, and provide a precise rate-level characterization of the tradeoff between node privacy and SBM recovery.

In cellular networks, base stations broadcast configurations that devices use for the random access procedure, which is a vital part of the connection setup. Ideally, the network should choose configurations based on the deployment scenario to optimize radio resource management. Doing so can, for example, decrease collisions of random access messages. We captured 112,806 data points of cellular broadcast information from nine network operators across three countries and analyzed how the operators configure the random access procedure. We found that configurations often do not fit the deployment scenario, and neighboring cells often use the same configuration, causing an unnecessarily high risk of collisions and, hence, delay in the connection setup. Furthermore, we simulated the random access procedure in NS-3 and found that by varying the configurations in a large area with many cells, the number of collisions can be reduced by 43% on average and up to 61%, and the connection delay can be lowered by 11% on average and up to 42%. Our findings indicate that simple adaptations in the random access configurations can greatly improve the performance of cellular networks.

The data-driven linear quadratic regulator (ddLQR) is a widely studied control method for unknown dynamical systems with disturbance. Existing approaches, both indirect, i.e., those that identify a model followed by model-based design, and direct, which bypasses the identification step, often rely on the certainty-equivalence principle and therefore do not explicitly account for model uncertainty. In this paper, we propose a Bayesian formulation for both indirect and direct ddLQR that incorporates posterior uncertainty into the control design. The resulting expected cost decomposes into a certainty-equivalence term and a variance-dependent term, providing a principled interpretation of regularization. We further show that the indirect and direct formulations are equivalent under this perspective. The resulting direct method admits a tractable semidefinite program whose size is independent of the data length. Numerical simulations demonstrate improved optimality gap and closed-loop stability, particularly in low-data regimes.

Diffusion model deployment has been suffering from high energy consumption and inference latency despite its superior performance in visual generation tasks. Dynamic voltage and frequency scaling (DVFS) offers a promising solution to exploit the potential of the underlying accelerators. However, existing approaches often lead to either limited efficiency gains or degraded output quality because they overlook the inherent fault tolerance of the diffusion model. Therefore, in this paper, we propose DRIFT, a novel algorithmarchitecture co-optimization framework that harnesses the fault tolerance for efficient and reliable diffusion model inference. We first perform a comprehensive resilience analysis on representative diffusion models. Building on these observations, we introduce a fine-grained, resilience-aware DVFS strategy that selectively protects error-sensitive network blocks and timesteps, and a rollback algorithm-based fault tolerance (ABFT) mechanism that adaptively corrects only critical errors by reverting to previous timesteps. We further optimize offloading intervals and reorganize data layouts to reduce memory overhead. Experiments across diverse models and datasets show that DRIFT can achieve on average 36% energy savings through voltage underscaling or 1.7x speedup via overclocking while maintaining generation quality.

We present a numerical program that evaluates deep inelastic scattering (DIS) structure functions at next-to-leading order (NLO) accuracy in the dipole picture. In this numerical implementation the NLO DIS impact factors with massive quarks are written in a form that ensures a stable numerical evaluation of the DIS cross sections.

Beamforming is conventionally understood as a collective property of many discrete antenna elements in both communication and radar fields, which links angular selectivity to array size, element spacing, and band-specific hardware. Here we uncover a fundamentally different beamforming mechanism achieved by a Rydberg atomic receiver: a Rydberg-atom vapor cell dressed by a local-oscillator field constitutes a continuous quantum aperture. In this regime, spatially-varying quantum coherence across the aperture provides continuous amplitude-phase control, allowing a directional beam pattern to emerge from one sensing volume rather than from an engineered array. We establish the theory of continuous quantum aperture and show that tailoring the local-oscillator field can directly program the aperture response. This enables reconfigurable single-peak, multipeak, and multiband beamforming within a single vapor cell. Experiments on a Rydberg atomic receiver prototype verify that practical beam patterns agree with theoretical predictions across aperture sizes, frequency bands, and local-oscillator configurations. Leveraging this new beamforming mechanism, we further demonstrate interference mitigation, multiuser access, and multiband multiuser access with the single-vapor-cell platform. Our results identify the continuous quantum aperture as a new operating principle of Rydberg atomic receivers and establish single-vapor-cell beamforming as an integrated and reconfigurable platform for spatially selective electromagnetic reception.

A circular pursuit guidance problem involving pursuer-target engagement is studied in this paper using a bifurcation theory based numerical approach. While target is modeled as a point mass moving around in a circle with certain velocity, pursuer dynamics is driven by the relative position and orientation with respect to the target. A planar case is currently considered. A mathematical model representing the engagement scenario is derived and two cases are presented, one without and the other with a basic model for pursuer speed dynamics accounting for limitations imposed by available force. Analytical and simulation results are presented to elucidate the novel approach. Advantages of using this approach for arriving at laws for pursuer-target engagement are highlighted.

Bistatic backscatter communication requires strong illumination of a backscatter device (BD), while a spatially separated reader detects the weak modulated reflection. In practice, the resulting direct link interference (DLI) at the reader can dominate the received backscattered signal and limit detection performance. This paper experimentally investigates transmit beamforming that jointly maximizes BD illumination and suppresses DLI at the reader in a distributed multiple-input multiple-output setup. We compare phase-only maximum ratio transmission (PO-MRT) with the proposed direct-link suppression (DLS) scheme, which enforces a spatial null at the reader under per-antenna power constraints. Measurements using a phase-coherent 42-element ceiling array at 920 MHz show that DLS reduces the DLI at the target reader and improves the signal-to-interference ratio by up to 31 dB compared to PO-MRT.

With the rapid adoption of large language models (LLMs) in financial service scenarios, dialogue security detection under high regulatory risk presents significant challenges. Existing methods mainly rely on single-dimensional semantic judgments or fixed rules, making them inadequate for handling multi-turn semantic evolution and complex regulatory clauses; moreover, they lack models specifically designed for financial security detection. To address these issues, this paper proposes FinSec, a four-tier security detection framework for financial agent. FinSec enables structured, interpretable, and end-to-end identification of actual financial risks, incorporating suspicious behavior pattern analysis, delayed risk and adversarial inference, semantic security analysis, and integrated risk-based decision-making. Notably, FinSec significantly enhances the robustness of high-risk dialogue detection while maintaining model utility. Experimental results demonstrate FinSec's leading performance. In terms of overall detection capability, FinSec achieves an F1 score of 90.13%, improving upon baseline models by 6--14 percentage points; its ASR is reduced to 9.09%, markedly lowering the probability of unsafe outputs; and the AUPRC increases to 0.9189 -- an approximate 9.7% gain over general frameworks. Additionally, in balancing utility and safety, FinSec obtains a composite score of 0.9098, delivering robust and efficient protection for financial agent dialogues.

Constructing valid inferential methods for constrained parameters in normal and Poisson distributions represents two fundamental and important problems in applied statistics, for which there is currently no unified framework for statistical inference. Most existing studies assume that the nuisance parameters of the model are known, an assumption that is often impractical in real-world applications. However, under the more realistic scenario where nuisance parameters are unknown, the available Bayesian interval estimation methods fail to guarantee nominal coverage and thus cannot provide exact inference. To address these limitations, this paper develops prior-free inferential model (IM) approaches for parameters of interest in constrained normal and Poisson models and demonstrates that the confidence intervals (CIs) obtained from these novel IM methods can achieve exact nominal coverage. Furthermore, considering the discrete nature of the Poisson distribution, we employ random weighting techniques to improve the conservative coverage performance of the IM CIs. Simulation studies show that the coverage probabilities of the improved nonrandomized inferential model (NIM) CIs are closest to the prespecified nominal levels, with corresponding expected lengths shorter than those of Bayesian intervals in weak signal scenarios, whereas the shorter expected lengths of Bayesian intervals in strong signal scenarios come at the cost of sacrificing coverage guarantees. Therefore, the proposed IM and NIM CIs are superior to the Bayesian CIs. Finally, the advantages of the proposed methods are confirmed through an analysis of two experimental datasets on neutrinos in high-energy physics.

This paper provides an introduction to the theory of semiconductor laser diodes, with special focus on their noise properties. It may be considered an additional chapter to the textbook [1]. As such, it will also refer to equations in that book.

Superconducting coplanar waveguide resonators are essential components in quantum processors, where their internal quality factor (Qi) constrains qubit coherence and readout fidelity. In niobium devices, microwave losses at millikelvin temperatures are strongly influenced by two-level systems (TLS) associated with the complex NbOx surface oxide. To mitigate these losses, we investigate a surface-engineering approach in which Nb films are capped in situ with a thin tantalum layer to suppress Nb2O5 formation and replace the native NbOx interface with a Ta-based oxide. We fabricate Nb/Ta bilayer and reference Nb resonators on high-resistivity silicon using identical DC sputtering and wet etching conditions, and characterize their performance at millikelvin temperatures. Fresh Ta-encapsulated devices exhibit internal quality factors up to 2.4 x 10^6 in the near-single-photon regime, with power dependence consistent with reduced TLS-related loss at the metal-air interface. A control Nb device fabricated under the same process shows comparatively lower Q_TLS, consistent with the beneficial effect of the Ta capping layer. Furthermore, ageing tests performed on Nb/Ta resonators after six months reveal a moderate reduction in Q_TLS relative to their initial values, yet the performance remains superior to newly fabricated Nb-only devices. These results suggest that thin Ta encapsulation enhances interface quality and contributes to improved temporal stability while remaining compatible with Nb-based fabrication workflows.

With the Event Horizon Telescope and future Very Long Baseline Interferometry arrays poised to image supermassive black holes, there is an urgent need to understand dynamic aspects of small-scale structure near the supermassive black hole. In this study, we introduce the relative magnification factor to characterize point sources distributed on the surface of the accretion disk near a black hole. We investigate the influence of source motion on this factor, comparing static sources with those corotating with the disk. In contrast to the static case, which can be well-understood in the standard framework of gravitational lensing, corotating sources exhibit significant distortions in the distribution of the magnification factor on both the image and source planes, indicating that the caustic structure is substantially modulated by source motion. This magnification factor pattern encodes signatures of the kinematics of accretion flow when the time-delayed images are incorporated. This potentially offers a novel probe for investigating the interplay between spacetime geometry and properties of accretion flow.

In this paper, we obtain some important inequalities for a class of Hessian quotient type operators $\frac{σ_k(Λ(D^2u))}{σ_l(Λ(D^2u))}$, which can be regarded as a generalization of the classical Hessian quotient operators. As an application, we establish global a priori estimates and prove an existence theorem for the Neumann problem of the corresponding degenerate Hessian quotient type equation, in which the admissible range of $k$ is extended to $0< k \leq C^\mathbf{p}_n$ with $1 \leq \mathbf{p} \leq n-1$.

The origin of low surface brightness (LSB) galaxies remains a key open question in galaxy formation, reflecting the balance internal mechanisms and environmental influence. Using MaNGA integral-field spectroscopy, we investigate whether LSB and high surface brightness (HSB) galaxies of comparable stellar mass ($9 < \log M_\ast < 10$) occupy distinct environments or differ primarily through internal evolution. Our late-type sample comprises 113 central and 29 satellite LSB galaxies, and 374 central and 142 satellite HSB galaxies. We characterize environments on scales from 100 kpc to 10 Mpc, analyzing radial profiles of stellar mass surface density ($Σ_\ast$), star formation activity, and gas-phase metallicity. Central LSB and HSB galaxies inhabit similarly low-density large-scale ($>$200 kpc) environments, but LSB galaxies are more isolated on small scales ($\sim$100 kpc). Even after matching in stellar mass and environment, LSB galaxies show systematically lower $Σ_\ast$, $Σ_{SFR}$, and metallicities, often hosting diffuse, weakly star-forming bulges embedded in extended disks. These results indicate that LSB structure and star formation are not primarily governed by large-scale environment or halo mass. While secondary halo properties such as spin, concentration, or gas accretion history are often invoked, their environmental dependence appears weak. Instead, LSB-HSB differences for centrals likely reflect divergent assembly or interaction histories and internal processes -- such as angular momentum-driven disk evolution or inefficient gas conversion -- largely decoupled from large-scale environment. Nonetheless, environment still influences the observed star formation and chemical differences between central and satellite LSB galaxies.

The disk instability (DI) model for giant planet formation remains an attractive alternative in explaining the formation of giant planets at early times, giant planets at large radial distances, and giant planets orbiting M-stars. In this review, we present recent developments in the disk instability model including hydrodynamical as well as magneto-hydrodynamical (MHD) disk simulations, populations synthesis models, and simulations of clump-clump collisions. We also discuss advances in observations that can be used to constrain and test this formation scenario.

Giannelli and Chiribella derived an observable-generator duality for energy from a collision model of informational nonequilibrium. We study a continuous-variable version aimed at the Galilean one-particle sector. A smooth family of reference states around an isotropic equilibrium supplies time, translation, rotation, and boost directions. The local observable-generator correspondence is obtained by differentiating a smooth extension of the single-state duality map, and the norm-one property of localization is obtained from a fiducial focusing assumption together with covariance. Combined with the standard smearing form of covariant localization observables, this yields sharp localization. With local inertial composition, the spin-cover action of rotations, and a central boost-translation holonomy, every irreducible sector is unitarily equivalent to the Hilbert space L2(R3) tensored with a (2s+1)-dimensional spin space. In that representation translations are generated by the canonical momentum, the holonomy is a scalar mass m > 0, boosts at t = 0 are generated by m times the position observable, the Hamiltonian is the free-particle kinetic term plus a constant E0, and the total angular momentum is orbital plus spin.

Acquiring the channel state information from limited and noisy observations at pilot positions is critical for wireless multiple-input multiple-output (MIMO)-orthogonal frequency division multiplexing (OFDM) systems. In this paper, we view this process as a conditional generative task in which the partial noisy channel estimates at the pilots are utilized as a ``prompt'' to guide the diffusion ``inpainting'' of the underlying channel. To this end, we resort to a general Conditional Diffusion Transformer (CDiT) framework with a well-designed network architecture and update rule. In particular, we design a dedicated embedding strategy to encode and adapt to different pilot patterns and noise levels, and utilize a special cross-attention mechanism to align the partial raw channel observations with the denoised channel at each time step of the generation process. This architecture effectively anchors the diffusion process, enabling the model to accurately recover full channel details from limited noisy observations. Comprehensive experimental results show that, the proposed approach achieves a performance gain of over 5 dB compared to the baselines under varying noise conditions, and provides robust channel acquisition even under a sparse pilot density of 1/32 without significant performance loss compared to the denser pilot cases. Moreover, it is capable of generating high-quality channel matrices within just 10 inference steps, effectively balancing estimation accuracy with computational efficiency and inference speed. Ablation studies demonstrate the rationality of the model design and the necessity of its modules.

In this paper we examine a multivariate risk model, with common renewal counting process, constant interest rate, and each claim vector is accompanied by a random number of delayed claim vectors. The interest is focused on the asymptotic behavior of the entrance probability of the discounted aggregate claims into some rare-sets, over a finite and an infinite time horizon. Our results study the the case where the main claims and the delayed claims have in some sense, asymptotic equivalent tails, but also the case where the delayed claims are negligible with comparisons with the main claims. More precisely, our estimations over finite time horizon are equipped with local uniformity, and are valid under the assumption of multivariate subexponential distributions for the claim distributions. On the case of infinite time horizon we need a mild restriction on the distribution class of multivariate subexponential distributions with positive lower Karamata index. The asymptotic relations reflect completely as all the sources of randomness, under the concrete rare-sets A, and the different dependence structures as well, without loosing elegance in spite of their generality. Further, we provide some more explicit formulas, together with relaxations of some assumptions, for the claim distributions from the multivariate regular variation. For the proof of the main results on infinite time case and for the construction of examples of multivariate distributions we need some closure properties of subexponential distributions with positive lower Karamata index. Especially, we present some necessary and sufficient conditions for the closure property with respect to convolution and some sufficient conditions for the closure property with respect to product convolution. Finally, we carry out some numerical studies to show the accuracy of our asymptotic estimations.