arXiv subjects

Ziyi Li

Publications and source records attributed to Ziyi Li.

At least 19 recordsLinked to original sources

arXiv2026

SurvFM enables tabular foundation models for right-censored survival prediction

General-purpose tabular foundation models can be adapted across prediction tasks, but right censoring leaves many event times unknown and prevents their direct use as regression labels. SurvFM converts censored follow-up into observation-level targets for restricted mean survival time (RMST), the expected event-free time accumulated up to a chosen horizon. These targets allow multiple tabular foundation models to predict RMST without architectural modification. In simulations with known RMST, SurvFM achieved leading RMST accuracy and competitive discrimination across heterogeneous settings. Its targets were more accurate than simpler outcome constructions, with larger gains as censoring increased. Across 55 public datasets, SurvFM models remained in the leading performance band under full and restricted training. Models fitted in either of two public myelodysplastic syndrome cohorts retained competitive performance in the other without refitting. SurvFM separates censoring handling from prediction architecture, allowing advances in general-purpose tabular prediction to enter survival analysis without model-specific redesign.

stat.ME↗

arXiv2026

Probing for Better Age of Information in Energy-Harvesting Random Access Networks

In this paper, we investigate the impact of channel probing and reservation on the Age of Information (AoI) in energy-harvesting (EH) random access networks, where each source relies solely on harvested energy for status updating. To mitigate collisions, each node may expend a small amount of energy to send a probing signal before transmission, and a successful probe reserves the channel in the current slot. If probing fails, the node can either remain silent, termed strict avoid free competition (SAFC), attempt data transmission with a certain probability, termed reserved nodes competition (RUC), or adopt all-active nodes competition (AUC), where all energy-sufficient nodes may contend regardless of whether they probed. We derive closed-form expressions for the network-average AoI under these three schemes and validate them via simulations. The results show that AUC consistently achieves the lowest AoI by shortening the waiting time to convert harvested energy into successful updates. This finding challenges the conventional wisdom that strict collision avoidance is always optimal in energy-constrained systems, since allowing additional contention can effectively amortize probing overhead across more transmission opportunities. Comparisons with EH-enabled slotted ALOHA further show that probing-based access significantly outperforms direct transmission in energy-constrained regimes, highlighting channel probing as an effective approach to improving freshness.

cs.NI↗

arXiv2026

Spinning States and Unitarity in 3D Gravity

We revisit the proposal to cure the negative density of states in the three-dimensional gravitational path integral by adding spinning states whose spin scales with the central charge. We show that sub-extremal and extremal spinning states below the black hole threshold, together with certain overspinning states above it, can cancel the known negativities in the spectrum. We interpret the states below threshold as bulk spinning defects. The overspinning states were previously interpreted as classical spinning strings; here we instead study their realization as maximal, fixed-point-free BTZ quotients. All of these spinning geometries exhibit causal pathologies in their Lorentzian continuations. The overspinning geometries arise from mixed elliptic--hyperbolic identifications and possess one formally real chiral temperature. We derive the extremal scalar correlator and propose a quotient-motivated image expression for the overspinning background.

hep-th↗

arXiv2026

Tell2Adapt: A Unified Framework for Source Free Unsupervised Domain Adaptation via Vision Foundation Model

Source Free Unsupervised Domain Adaptation (SFUDA) is critical for deploying deep learning models across diverse clinical settings. However, existing methods are typically designed for low-gap, specific domain shifts and cannot generalize into a unified, multi-modalities, and multi-target framework, which presents a major barrier to real-world application. To overcome this issue, we introduce Tell2Adapt, a novel SFUDA framework that harnesses the vast, generalizable knowledge of the Vision Foundation Model (VFM). Our approach ensures high-fidelity VFM prompts through Context-Aware Prompts Regularization (CAPR), which robustly translates varied text prompts into canonical instructions. This enables the generation of high-quality pseudo-labels for efficiently adapting the lightweight student model to target domain. To guarantee clinical reliability, the framework incorporates Visual Plausibility Refinement (VPR), which leverages the VFM's anatomical knowledge to re-ground the adapted model's predictions in target image's low-level visual features, effectively removing noise and false positives. We conduct one of the most extensive SFUDA evaluations to date, validating our framework across 10 domain adaptation directions and 22 anatomical targets, including brain, cardiac, polyp, and abdominal targets. Our results demonstrate that Tell2Adapt consistently outperforms existing approaches, achieving SOTA for a unified SFUDA framework in medical image segmentation. Code are avaliable at https://github.com/derekshiii/Tell2Adapt.

cs.CV↗

arXiv2026

GraFSTNet: Graph-based Frequency SpatioTemporal Network for Cellular Traffic Prediction

With rapid expansion of cellular networks and the proliferation of mobile devices, cellular traffic data exhibits complex temporal dynamics and spatial correlations, posing challenges to accurate traffic prediction. Previous methods often focus predominantly on temporal modeling or depend on predefined spatial topologies, which limits their ability to jointly model spatio-temporal dependencies and effectively capture periodic patterns in cellular traffic. To address these issues, we propose a cellular traffic prediction framework that integrates spatio-temporal modeling with time-frequency analysis. First, we construct a spatial modeling branch to capture inter-cell dependencies through an attention mechanism, minimizing the reliance on predefined topological structures. Second, we build a time-frequency modeling branch to enhance the representation of periodic patterns. Furthermore, we introduce an adaptive-scale LogCosh loss function, which adjusts the error penalty based on traffic magnitude, preventing large errors from dominating the training process and helping the model maintain relatively stable prediction accuracy across different traffic intensities. Experiments on three open-sourced datasets demonstrate that the proposed method achieves prediction performance superior to state-of-the-art approaches.

cs.NI↗

arXiv2025

Sark: Oblivious Integrity Without Global State

In this paper, we introduce Sark, a reference architecture for transferring unforgeable, stateful, oblivious (USO) assets. We describe the motivation, design, and implementation of the core subsystems of Sark, Porters, which accumulate and roll-up commitments from Clients, and Sloop, a permissioned, crash fault-tolerant (CFT) blockchain system. We analyse the operation of the system using STRIDE threat analysis, and the `CIA Triad': Confidentiality, Availability, and Integrity. We then introduce the concept of \textit{local centrality} and use it to address design trade-offs related to decentralization. Finally, we point to future work on Byzantine fault-tolerance (BFT), and mitigating the local centrality of Porters.

cs.CR↗

arXiv2025

Spinning Fields in Lorentzian AdS

We construct the higher spin wave functions in the embedding space of anti-de Sitter Lorentzian spacetime. These wave functions are built from a primary wave functions that has a simple structure expressed in terms of the special conformal generator vector fields in AdS. We compute the eigenvalue of the quadratic Casimir for the symmetric traceless states, and show explicitly that these satisfy the higher spin wave equation. We also demonstrate that these wave functions have the right structure in the flat space limit for massive higher spin fields, and can be used to construct $in$ and $out$ states for scattering processes. Spinning states that become massless in the flat limit are extremely subtle. The problem can be isolated to longitudinal polarizations.

hep-th↗

arXiv2025

Boosting Neural Video Representation via Online Structural Reparameterization

Neural Video Representation~(NVR) is a promising paradigm for video compression, showing great potential in improving video storage and transmission efficiency. While recent advances have made efforts in architectural refinements to improve representational capability, these methods typically involve complex designs, which may incur increased computational overhead and lack the flexibility to integrate into other frameworks. Moreover, the inherent limitation in model capacity restricts the expressiveness of NVR networks, resulting in a performance bottleneck. To overcome these limitations, we propose Online-RepNeRV, a NVR framework based on online structural reparameterization. Specifically, we propose a universal reparameterization block named ERB, which incorporates multiple parallel convolutional paths to enhance the model capacity. To mitigate the overhead, an online reparameterization strategy is adopted to dynamically fuse the parameters during training, and the multi-branch structure is equivalently converted into a single-branch structure after training. As a result, the additional computational and parameter complexity is confined to the encoding stage, without affecting the decoding efficiency. Extensive experiments on mainstream video datasets demonstrate that our method achieves an average PSNR gain of 0.37-2.7 dB over baseline methods, while maintaining comparable training time and decoding speed.

eess.IV↗

arXiv2025

Radial oscillations of neutron stars in Starobinsky gravity and its Gauss-Bonnet extension

Starobinsky gravity, as one of the simplest and best-behaved higher-curvature gravity theories, has been extensively studied in the context of neutron stars over the past few decades. In this work, we investigate the adiabatic radial oscillation stability of neutron stars within the framework of Starobinsky gravity. We find that gravitational modifications can significantly impact stellar stability. Specifically, the higher-derivative nature of the theory causes the exterior spacetime to dynamically respond to fluid oscillations, in contrast to general relativity where Birkhoff's theorem ensures a static exterior. For stellar models with low central densities, the fundamental frequency becomes nearly independent of the central density when the coupling constant is large. For stellar models with high central densities, the transition from stability to instability still approximately occurs near the maximum-mass configuration, similar to the case in general relativity. Our main analysis is conducted in the Jordan frame of the scalar-tensor gravity equivalent to Starobinsky gravity, and we explicitly verify consistency with results obtained in the Einstein frame. We further extend our study to a class of Gauss-Bonnet extensions of Starobinsky gravity.

gr-qc↗

arXiv2025

A hybrid two-level weighted Schwartz method for time-harmonic Maxwell equations

This paper concerns the preconditioning technique for discrete systems arising from time-harmonic Maxwell equations with absorptions, where the discrete systems are generated by N\'ed\'elec finite element methods of fixed order on meshes with suitable size. This kind of preconditioner is defined as a two-level hybrid form, which falls into the class of ``unsymmetrically weighted'' Schwarz method based on the overlapping domain decomposition with impedance boundary subproblems. The coarse space in this preconditioner is constructed by specific eigenfunctions solving a series of generalized eigenvalue problems in the local discrete Maxwell-harmonic spaces according to a user-defined tolerance $\rho$. We establish a stability result for the considered discrete variational problem. Using this discrete stability, we prove that the two-level hybrid Schwarz preconditioner is robust in the sense that the convergence rate of GMRES is independent of the mesh size, the subdomain size and the wave-number when $\rho$ is chosen appropriately. We also define an economical variant that avoids solving generalized eigenvalue problems. Numerical experiments confirm the theoretical results and illustrate the efficiency of the preconditioners.

math.NA↗

arXiv2024

Neutron stars in Gauss-Bonnet extended Starobinsky gravity

Recently, a class of Gauss-Bonnet extended Starobinsky gravity was proposed, allowing black holes to carry ghost-free massive scalar hair for the first time without requiring additional matter fields. This intriguing feature offers a new perspective for understanding higher-curvature pure gravity and highlights the importance of further studying the potential effects of Gauss-Bonnet extensions in gravitational systems beyond black holes. In this study, we investigate the properties of neutron stars within this model, focusing on how the higher-curvature terms, particularly the coupling between the Gauss-Bonnet term and the curvature-squared term, impact the stellar structure. We present a detailed analysis of these effects and compute the moment of inertia for rotating neutron stars under the slow-rotation approximation. The substantial differences in the moment of inertia between general relativity and Gauss-Bonnet extended Starobinsky gravity are expected to be detectable through future high-precision observations.

gr-qc↗

arXiv2024

A Meta-analysis of College Students' Intention to Use Generative Artificial Intelligence

It is of critical importance to analyse the factors influencing college students' intention to use generative artificial intelligence (GenAI) to understand and predict learners' learning behaviours and academic outcomes. Nevertheless, a lack of congruity has been shown in extant research results. This study, therefore, conducted a meta-analysis of 27 empirical studies under an integrated theoretical framework, including 87 effect sizes of independent research and 33,833 sample data. The results revealed that the main variables are strongly correlated with students' behavioural intention to use GenAI. Among them, performance expectancy (r = 0.389) and attitudes (r = 0.576) play particularly critical roles, and effort expectancy and habit are moderated by locational factors. Gender, notably, only moderated attitudes on students' behavioural intention to use GenAI. This study provides valuable insights for addressing the debate regarding students' intention to use GenAI in existed research, improving educational technology, as well as offering support for school decision-makers and educators to apply GenAI in school settings.

cs.CY↗

arXiv2024

Aspects of Holography in Conical $AdS_3$

We study the Feynman propagator of free scalar fields in $AdS_3$ with a conical defect. The propagator is built by solving the bulk equation of motion, summing over the modes of the field, and taking the boundary limit. We then perform several consistency checks. In the dual CFT, the operator responsible for the defect creates a highly excited state. We consider the exchange of the Virasoro identity block in the heavy-light limit to obtain an expression for the propagator sensitive to the mass of the defect. In $AdS_3/\mathbb{Z}_n$, we treat the propagator by the method of images and in the geodesic approximation. More generally, we argue that long-range correlations of the scalar are suppressed as the defect becomes more massive: we find a continuous phase transition in the correlator at the BTZ threshold and examine its critical behavior. Finally, we apply our results to holographic entanglement entropy using an analogy between our scalars and replica twist fields.

hep-th↗

arXiv2024

Spinning Particle Geometries in AdS$_3$/CFT$_2$

We study spinning particle/defect geometries in the context of AdS$_3$/CFT$_2$. These solutions lie below the BTZ threshold, and can be obtained from identifications of AdS$_3$. We construct the Feynman propagator by solving the bulk equation of motion in the spinning particle geometry, summing over the modes of the fields and passing to the boundary. The quantization of the scalar fields becomes challenging when confined to the regions that are causally well-behaved. If the region containing closed timelike curves (CTCs) is included, the normalization of the scalar fields enjoys an analytical simplification and the propagator can be expressed as an infinite sum over image geodesics. In the dual CFT$_2$, the propagator can be recast as the HHLL four-point function, where by taking into account the $PSL (2,\mathbb{Z})$ modular images, we recover the bulk computation. We comment on the casual behavior of bulk geometries associated with single-trace operators of spin scaling with the central charge below the BTZ threshold.

hep-th↗

arXiv2024

A novel coarse space applying to the weighted Schwarz method for Helmholtz equations

In this paper we are concerned with restricted additive Schwarz with local impedance transformation conditions for a family of Helmholtz problems in two dimensions. These problems are discretized by the finite element method with conforming nodal finite elements. We design and analyze a new adaptive coarse space for this kind of restricted additive Schwarz method. This coarse space is spanned by some eigenvalue functions of local generalized eigenvalue problems, which are defined by weighted positive semi-definite bilinear forms on subspaces consisting of local discrete Helmholtz-harmonic functions from impedance boundary data. We proved that a two-level hybrid Schwarz preconditioner with the proposed coarse space possesses uniformly convergence independent of the mesh size, the subdomain size and the wave numbers under suitable assumptions. We also introduce an economic coarse space to avoid solving generalized eigenvalue problems. Numerical experiments confirm the theoretical results.

math.NA↗

arXiv2023

CEDNet: A Cascade Encoder-Decoder Network for Dense Prediction

Multi-scale features are essential for dense prediction tasks, such as object detection, instance segmentation, and semantic segmentation. The prevailing methods usually utilize a classification backbone to extract multi-scale features and then fuse these features using a lightweight module (e.g., the fusion module in FPN and BiFPN, two typical object detection methods). However, as these methods allocate most computational resources to the classification backbone, the multi-scale feature fusion in these methods is delayed, which may lead to inadequate feature fusion. While some methods perform feature fusion from early stages, they either fail to fully leverage high-level features to guide low-level feature learning or have complex structures, resulting in sub-optimal performance. We propose a streamlined cascade encoder-decoder network, dubbed CEDNet, tailored for dense \mbox{prediction} tasks. All stages in CEDNet share the same encoder-decoder structure and perform multi-scale feature fusion within the decoder. A hallmark of CEDNet is its ability to incorporate high-level features from early stages to guide low-level feature learning in subsequent stages, thereby enhancing the effectiveness of multi-scale feature fusion. We explored three well-known encoder-decoder structures: Hourglass, UNet, and FPN. When integrated into CEDNet, they performed much better than traditional methods that use a pre-designed classification backbone combined with a lightweight fusion module. Extensive experiments on object detection, instance segmentation, and semantic segmentation demonstrated the effectiveness of our method. The code is available at https://github.com/zhanggang001/CEDNet.

cs.CV↗

arXiv2023

Open-vocabulary Object Segmentation with Diffusion Models

The goal of this paper is to extract the visual-language correspondence from a pre-trained text-to-image diffusion model, in the form of segmentation map, i.e., simultaneously generating images and segmentation masks for the corresponding visual entities described in the text prompt. We make the following contributions: (i) we pair the existing Stable Diffusion model with a novel grounding module, that can be trained to align the visual and textual embedding space of the diffusion model with only a small number of object categories; (ii) we establish an automatic pipeline for constructing a dataset, that consists of {image, segmentation mask, text prompt} triplets, to train the proposed grounding module; (iii) we evaluate the performance of open-vocabulary grounding on images generated from the text-to-image diffusion model and show that the module can well segment the objects of categories beyond seen ones at training time; (iv) we adopt the augmented diffusion model to build a synthetic semantic segmentation dataset, and show that, training a standard segmentation model on such dataset demonstrates competitive performance on the zero-shot segmentation(ZS3) benchmark, which opens up new opportunities for adopting the powerful diffusion model for discriminative tasks.

cs.CV↗

arXiv2023

Post-Quantum $κ$-to-1 Trapdoor Claw-free Functions from Extrapolated Dihedral Cosets

\emph{Noisy trapdoor claw-free function} (NTCF) as a powerful post-quantum cryptographic tool can efficiently constrain actions of untrusted quantum devices. However, the original NTCF is essentially \emph{2-to-1} one-way function (NTCF$^1_2$). In this work, we attempt to further extend the NTCF$^1_2$ to achieve \emph{many-to-one} trapdoor claw-free functions with polynomial bounded preimage size. Specifically, we focus on a significant extrapolation of NTCF$^1_2$ by drawing on extrapolated dihedral cosets, thereby giving a model of NTCF$^1_κ$ where $κ$ is a polynomial integer. Then, we present an efficient construction of NTCF$^1_κ$ assuming \emph{quantum hardness of the learning with errors (LWE)} problem. We point out that NTCF can be used to bridge the LWE and the dihedral coset problem (DCP). By leveraging NTCF$^1_2$ (resp. NTCF$^1_κ$), our work reveals a new quantum reduction path from the LWE problem to the DCP (resp. extrapolated DCP). Finally, we demonstrate the NTCF$^1_κ$ can naturally be reduced to the NTCF$^1_2$, thereby achieving the same application for proving the quantumness.

cs.CR↗