Searcharxiv⌕ Search

arXiv subjects

Hui Tian

Publications and source records attributed to Hui Tian.

At least 19 recordsLinked to original sources

Squeezing Drives Southward Magnetic Field Enhancement During 10 May 2024 CME-CME Interaction

The southward magnetic field carried by coronal mass ejections (CMEs) is a key driver of space weather events, yet how interacting CMEs amplify this field remains debated. The 10 May 2024 event, which produced the strongest geomagnetic storm since 2003, offers an ideal testbed for this question. Combining multi-spacecraft observations with high-resolution controlled numerical experiments, we track the CMEs continuously from their origin near the solar surface to beyond Earth, and find that inherent squeezing during CME-CME interactions generates the sustained enhancement of the southward field, exceeding 60 nT in magnitude near Earth. First driven by the dynamic pressure of the trailing CME and later maintained by magnetic and thermal pressures at the CME-CME interface, squeezing transforms the preceding CME from a bulky cloud into a flattened "pancake" structure, prevents expansion-induced field decay, and maintains an amplified southward field over long distances. The strength of this enhancement scales with the intensity of the squeezing. Our findings establish a direct link between CME-CME interaction and strong southward fields, laying a physics-based foundation for understanding and forecasting the geoeffectiveness of interacting CMEs.

astro-ph.SR↗

Scheduling Coflows in Multi-Core OCS Networks with Performance Guarantee

The coflow abstraction captures application-level communication patterns and enables coordinated scheduling of parallel flows to reduce job completion times in distributed systems. Modern data center networks (DCNs) are employing multiple independent optical circuit switching (OCS) cores operating concurrently to meet the massive bandwidth demands of application jobs. However, existing coflow scheduling research primarily focuses on the single-core setting, while studies of multi-core fabrics have largely considered electrical packet switching (EPS) networks. To address this gap, this paper studies the coflow scheduling problem in multi-core OCS networks under the not-all-stop reconfiguration model, in which the reconfiguration of one circuit does not interrupt other circuits. The challenges stem from two aspects: (i) cross-core coupling induced by traffic assignment across heterogeneous cores; and (ii) per-core OCS scheduling constraints, namely port exclusivity and reconfiguration delay. We propose an approximation algorithm that jointly integrates cross-core flow assignment and per-core circuit scheduling to minimize the total weighted coflow completion time (CCT) and establish a provable worst-case performance guarantee. Furthermore, our algorithm framework can be applied to the multi-core EPS scenario with a corresponding approximation guarantee for packet-switched fabrics. Trace-driven simulations using real Facebook workloads demonstrate that our algorithm can reduce the total weighted CCT and tail CCT.

cs.DC↗

Approximation Algorithms for Dual-Memory Request Scheduling in Disaggregated LLM Inference

Disaggregated LLM inference executes the prefill and decode phases on separate resource pools, reducing interference between the two phases. However, this architecture introduces a coupled scheduling problem because a request that completes prefill retains its prompt KV cache in prefill memory until it is admitted to the decode pool. Prefill and decode admission decisions therefore jointly determine memory occupancy and request completion times. To the best of our knowledge, this problem lacks a formal formulation, and no approximation guarantees are known for request scheduling in disaggregated LLM inference. We formalize this setting as the dual-memory request scheduling (DMRS) problem, whose objective is to minimize the sum of request completion times, and prove that it is strongly NP-hard. We develop two deterministic polynomial-time approximation algorithms. The first applies to instances satisfying $α<1$ and $β<1$ and achieves an approximation ratio of $8+\frac{4}{1-α}+\frac{8}{1-β}$, where $α$ and $β$ are the maximum normalized prefill and peak decode memory demands among all requests, respectively. In particular, it yields a $32$-approximation when $α,β\leq 1/2$. The second algorithm applies to all feasible instances and achieves a uniform $36$-approximation. Finally, evaluations on two H100 GPUs using trace-derived offline workloads show that the proposed algorithms reduce the sum of request completion times relative to multiple vLLM deployment baselines.

cs.DS↗

Learning to Watermark Speech Synthesis Against Model-Driven Reconstruction

Modern TTS systems increasingly generate synthetic speech at scale for diverse users. This setting calls for content-level provenance that can verify the origin of released speech and attribute it to the requesting user, which generative watermarking can support by embedding multi-bit identifiers directly into synthesized speech. Once released, however, speech may undergo heterogeneous learned transformations during distribution and editing, with reconstruction objectives that can preserve speech utility while affecting watermark recoverability differently. We find that no single watermark carrier remains consistently reliable across reconstruction models, as its survival depends jointly on the embedded structure, reconstruction mechanism, and observation representation. To this end, we propose Thrive, a multi-bit generative speech watermarking framework for modern autoregressive TTS, covering both discrete-token and continuous-representation generation under reconstruction attacks. Specifically, Rise synchronizes watermark injection into intermediate representations with its continued integration into subsequent generation, while Care combines waveform and spectral experts using bit-wise reliability selection. Experiments on both autoregressive paradigms show that Thrive preserves synthesis fidelity, achieves 87.6% average recovery accuracy under reconstruction attacks, and supports source attribution over candidate sets of up to 10,000 identities.

cs.CR↗

Robust Decentralized Federated Distillation via Multi-Modality Knowledge Collaboration

This paper propose a robust decentralized federated distillation method that enables clients with heterogeneous models to collaborate through predictions on shared unlabeled public data. In the proposed method, each client first evaluates the received predictions in three modalities of class prediction, boundary decision, and prediction correlation. It then filters unreliable clients, assigns reliability-based weights to the retained clients, and constructs a teacher for each type of knowledge. Finally, the corresponding distillation gradients are validated using a supervised gradient computed from private data. Conflicting prediction and boundary gradients are removed, and conflicting relation gradients are suppressed before the final model update. We prove the convergence of the proposed method by showing stable local optimization for honest clients under Byzantine distillation. Particularly, we show that our method ensures a bounded Byzantine influence on both distillation gradients and individual client private gradients after cross-modality fusion, thereby enabling stable local optimization for honest clienunder Byzantine distillation. Extensive experiments on CIFAR-10 and CIFAR-100 demonstrate that the proposed method improves the prediction accuracy of heterogeneous models of clients under non-IID data and Byzantine attacks. As the booming demands of federated learning in decentralized environments such as edge computing and mission-oriented UAV collaborations, our method has a great potential for adoption of DFL in unreliable real-world scenarios where clients are exposed to receiver-specific Byzantine messages of malicious predictions.

cs.LG↗

Robust Decentralized Personalized Federated Learning via Prediction-Constrained Neighborhood Collaboration

This paper proposes a robust decentralized personalized federated learning method R-DPFL, that enables clients to reduce the impact of Byzantine attacks via robust neighborhood direction estimation and history-based update trend prediction, rather than purely aggregating client models as in the existing work. In R-DPFL, each client first computes the current-round model update by aggregating the received neighborhood update vectors. It then predicts what this update should be based on its historical values and local model changes. Finally, R-DPFL computes the difference between these two quantities, adaptively clips this difference, and adds it to the local update. We prove convergence of the learning process through rigorous analysis and show that honest clients maintain stable personalized descent dynamics under Byzantine neighbor perturbations without requiring consensus among neighboring models. Extensive experiments on CIFAR-10 demonstrate that RDPFL consistently outperforms state-of-the-art decentralized and personalized federated learning baselines under heterogeneous and adversarial settings.

cs.LG↗

Solar Soft X-ray Coronal Dimming in a Failed Eruption Associated with Plasma Cooling

Coronal dimmings are observed as sudden and localized reductions in the extreme-ultraviolet and X-ray emission of the solar corona. Traditionally, significant dimmings of spectral lines formed at temperatures of 1-2 MK are regarded as indicators of coronal mass ejections (CMEs), reflecting the density depletion caused by plasma escaping into interplanetary space. In this Letter, we report a peculiar deep coronal dimming event predominantly observed in high-temperature spectral lines following an M8.8-class confined solar flare associated with a failed filament eruption. Sun-as-a-star measurements from the Geostationary Operational Environmental Satellite and the Extreme Ultraviolet Variability Experiment reveal intensity reductions exceeding 30% in soft X-ray (SXR) and measurable decreases in Fe XVIII (6.5 MK) and Fe XX (9.3 MK). Spatially resolved observations from the Atmospheric Imaging Assembly demonstrate that the dimming originates from the active region core, while the Solar Terrestrial Relations Observatory-A shows no evidence for CME-driven mass loss. Differential emission measure analysis reveals plasma at temperatures >5 MK cooling into lower temperatures, supporting plasma cooling as the dominant contributor to the hot-band dimming rather than CME-associated plasma escape. This event demonstrates that deep hot SXR dimmings can occur without substantial CME-driven mass loss, suggesting that alternative physical mechanisms may also account for unresolved stellar dimmings in addition to the commonly inferred CME signatures.

astro-ph.SR↗

AEGIS: Attention-Embedding Gradient Isolation Shield - Triple-Channel Gradient Masking for Privacy-Preserving Federated LLM Fine-Tuning

Gradient inversion attacks recover private training text from gradients shared in federated learning, posing a serious threat to collaborative model training. Through our analysis of transformer gradient structure, we identify three channels through which private token information leaks: the attention output projection gradient exposes a low-rank subspace that encodes input embeddings (Channel 1), the embedding gradient's row-norm sparsity directly reveals which tokens are present (Channel 2), and the MLP expansion gradient carries a recoverable subspace signal analogous to Channel 1 (Channel 3). State-of-the-art attacks exploit these channels analytically to achieve near-exact token recovery in seconds. Existing defences address at most one channel and either degrade model utility or leave the remaining structural signals intact. We introduce AEGIS (Attention-Embedding Gradient Isolation Shield), a lightweight defence that closes all three analytical channels with three backward-path operations requiring no architectural changes: freezing attention projection parameters eliminates Channel 1 by construction, calibrated noise injection into the embedding gradient destroys Channel 2's token-presence signal, and analogous per-block noise injection into the MLP expansion gradient masks Channel 3. The same masked gradient drives both the local optimiser step and the server export, so no clean signal is retained on either side. Evaluated across 11 models and six datasets, AEGIS reduces token recovery rates to near zero against a range of gradient inversion attacks, both analytical and optimisation-based, while preserving or improving model utility. We provide formal guarantees for Channels 1 and 2 and validate the full defence empirically against adaptive adversaries with complete knowledge of the mechanism.

cs.CR↗

Intense but Harmless: Exo-Space Weather Around an M Dwarf with a Single-Hemisphere Dynamo

M dwarfs are among the most promising host stars in the search for habitable exoplanets. However, their active atmospheres drive intense magnetic activity, including energetic flares and possibly coronal mass ejections (CMEs), which may pose serious threats to planetary habitability. In this study, we perform three-dimensional magnetohydrodynamic (MHD) simulations of CMEs on a fully convective M dwarf with a rotation period of 30 days, corresponding to the moderate-rotation regime. The magnetic topology driving our simulations is adopted from an exploratory global dynamo simulation of a fully convective low-mass star exhibiting a single-hemisphere magnetic configuration, which is not yet observationally confirmed. The large-scale magnetic field is mostly restricted to a single hemisphere and characterized by high-latitude polarity inversion lines (PILs), with the implication that most CMEs should originate from high latitudes. We find that these high-latitude CMEs propagate radially and away from the equatorial plane, producing only weak and spatially limited disturbances along the equatorial orbits of exoplanets. Moreover, low-latitude CMEs experience stronger drag within the dense and slow stellar wind near the equator, which significantly reduces both their propagation speeds and their overall impact on exoplanets. The resulting dynamic pressure enhancements on equatorial exoplanets caused by these CMEs are within two orders of magnitude above the quiescent conditions, much lower than those reported in previous M-dwarf CME simulations. These results indicate that, if such magnetic topologies indeed exist on M dwarfs, they may produce a relatively benign CME environment, which could be favorable for planetary habitability at face value.

astro-ph.SR↗

High-Frequency Magnetohydrodynamic Waves with Substantial Energy in the Solar Polar Corona

The acceleration and heating of the fast solar wind remain long-standing challenges in space physics. One type of leading theoretical models requires high-frequency magnetohydrodynamic (MHD) waves to transport and dissipate sufficient energy in the corona. However, such high-frequency waves with energetically significant amplitudes have never been unambiguously observed, leaving a key gap between theories and observations. Using high-cadence, high-resolution extreme-ultraviolet imaging from Solar Orbiter's Extreme Ultraviolet Imager, we identify a previously hidden population of high-frequency MHD waves in coronal plumes of the solar polar region. An analysis of the detected propagating kink waves shows that over one-third have periods shorter than 100 s, a population largely undetected by earlier instruments. Power spectral analysis demonstrates that these high-frequency waves carry substantial energy flux, which are significantly underestimated in lower-cadence data. These results suggest that high-frequency MHD waves may contribute importantly to the energy budget of the solar polar corona and could play a role in solar wind acceleration, highlighting the value of high-resolution observations for probing energy transport in magnetized space and astrophysical plasmas.

astro-ph.SR↗

Understanding the Travel-time Asymmetry of Acoustic Waves in Sunspots With Time-distance Helioseismology

Mapping the subsurface structure and flow field of sunspots has been a challenging task for helioseismology. In this work, we investigate the propagation of acoustic waves in a sunspot in NOAA active region 11312 using time-distance helioseismology. Travel times of waves traveling into and out of the sunspot are measured as functions of travel distance and azimuthal angle relative to the local radial direction. The same time-distance analysis is also applied to a simulated data based on a magnetohydrostatic (MHS) model of sunspot, and forward modeling of travel times is performed using ray tracing based on both the MHS sunspot model and a magnetohydrodynamic (MHD) simulation. We find that both ingoing (traveling from the quiet area into the sunspot) and outgoing waves (traveling from the sunspot into the quiet area) have shorter travel times than in the quiet Sun, with travel-time reductions of up to 40 s. The magnitude of the mean time shift is largest for waves traveling along the radial direction at small travel distances. A clear asymmetry is detected between ingoing and outgoing waves: outgoing waves generally exhibit shorter travel times. This asymmetry is strongest for radial direction and small travel distances, with differences exceeding 1 min for 3.5 mHz and 4.5 mHz waves. From the results of both observations and models, our analysis indicates that the overall reduction in travel time could be primarily caused by the combined effects of Wilson depression, magnetic field, and wave-speed perturbations, while the ingoing-outgoing asymmetry could be partly attributable to subsurface flows. Although the forward-modeling results reproduce several qualitative features of the observations, quantitative discrepancies remain, highlighting limitations of current sunspot models and ray-theoretical approximations.

astro-ph.SR↗

How Magnetic Field Strength Affects Stellar Coronal Mass Ejection Dynamics

Observations show that stellar coronal mass ejection (CME) candidates display relatively lower kinetic energies compared to expectations from solar flare-CME relations extrapolated to the stellar regime. This behaviour was predicted by studies of magnetic confinement of CMEs by strong large-scale stellar magnetic fields. However, the possible promoting role of stronger small-scale magnetic fields has not yet been properly explored in previous studies. In this work, we present the first parametric study that simultaneously incorporates both the promoting and confining effects of magnetic field strength on CME dynamics. We perform CME simulations with scaled solar magnetograms spanning = 1, 5, 10, 50, 100 B_sun and inserting flux ropes whose magnetic energy is set to scale as E_FR \propt ^2. Our results show that CME speed and mass increase with magnetic field strength in this restrictive scenario, approximately following v_CME \propt and M_CME \propt ^1.5. These trends indicate that, within this idealized solar-scaled framework, increasing the magnetic field strength enhances the net promoting forces relative to the confining forces and drives faster, more massive CMEs. We further identify the upward Lorentz force as the dominant contributor to the acceleration and the mass enhancement. We also conducted additional cases with different flux rope energies that do not follow the above scaling assumption, and found that stronger flux ropes produce faster and more massive CMEs for each given stellar model. The adopted scaling assumptions are intended as a controlled parametric experiment rather than a realistic model of young solar-type stars, and future work using more realistic stellar magnetic maps will be required to determine which regions of the parameter space explored here are most relevant to active stars.

astro-ph.SR↗

Fine-scale downflows above flare ribbons captured by Solar Orbiter/EUI

In solar flares, flare ribbons map chromospheric footpoints where flare energy deposition occurs. These locations are associated with field aligned energy transport from the corona that results from energy liberated during magnetic reconnection. Recent chromospheric observations in the H$α$ and H$β$ bands have revealed fine-scale downflow structures above flare ribbons, referred to as riblets. In this study, we identify similar downflow structures in the extreme-ultraviolet (EUV) wavelength using high-resolution observations from Solar Orbiter/EUI. These fine-scale downflows appear as downward-propagating, bright, and thread-like structures. They exhibit typical velocities of $\sim100~\mathrm{km\ s^{-1}}$, lifetimes of $\sim15$~s, and lengths of $\sim1.6$~Mm. Based on their morphological and dynamical properties, we interpret these observed downflows as the EUV counterparts of the riblets that have previously been reported from chromospheric observations. This study presents EUV imaging of $\sim 10^6$~K downflows above flare ribbons. We interpret these downflows as a result of (1) the energisation and subsequent compression of pre-existing chromospheric fibrils due to particle beams or (2) adiabatic or shock-driven compression induced by the downward-propagating plasma from the corona. These fine-scale EUV riblets provide a new diagnostic tool for probing the dynamics of magnetic reconnection as well as energy transport and deposition during solar flares.

astro-ph.SR↗

An O(K)-Approximation Coflow Scheduling in K-Core Optical Circuit Switching Networks

Coflow has emerged as a fundamental application-layer abstraction in distributed systems, enabling collaborative management of related flows to enhance job completion efficiency. To meet the increasing bandwidth demands of modern data center networks (DCNs), optical circuit switches are widely deployed due to their high capacity and energy efficiency. Simultaneously, DCN deployments are evolving towards heterogeneous parallel architectures, where multiple independent optical circuit switching (OCS) cores operate concurrently to facilitate bandwidth expansion and incremental upgrades. However, existing research on coflow scheduling in multi-core switching fabrics primarily focuses on electrical packet switching (EPS) networks, with a few known results on OCS networks without or with a poor performance guarantee. This paper studies the coflow scheduling problem in multi-core OCS networks under the not-all-stop reconfiguration model, focusing on two major challenges of overcoming cross-core coupling for inter-core traffic allocation and satisfying the constraints of port exclusivity and reconfiguration overhead for intra-core circuit scheduling. To minimize total weighted coflow completion time (CCT), we propose an efficient algorithm by integrating LP-guided global coflow ordering, inter-core flow allocation and intra-core circuit scheduling that achieves approximation ratios of $8K$ and $\left(8K+1\right)$ for zero and arbitrary release times of coflows, respectively, where $K$ is the number of OCS cores. This framework is also applicable to $H$-core EPS networks, providing approximation guarantees of $4H$ and $\left(4H+1\right)$ for zero-time and arbitrary-time release, respectively.

cs.DC↗

Neuroscience-inspired Staged Representation Learning with Disentangled Coarse- and Fine-Grained Semantics for EEG Visual Decoding

Decoding visual information from electroencephalography (EEG) signals remains a fundamental challenge in brain-computer interfaces and medical rehabilitation. Existing EEG visual decoding methods mainly focus on learning a single global EEG embedding for cross-modal alignment, but they largely overlook the staged and hierarchical characteristics of human visual processing. To address this limitation, we propose a neuroscience-inspired staged representation learning framework that reformulates EEG visual decoding as a stage-specific representation decomposition problem. The proposed framework organizes EEG representation learning into three complementary phases: low-level visual representation learning, high-level semantic representation learning, and integrative information fusion. To strengthen semantic modeling, we further introduce a multimodal dual-level semantic learning mechanism that separates coarse label-level semantics from fine image-level visual-semantic information. In addition, semantic latent channels are introduced as computational representation channels generated from observed visual EEG signals, expanding the channel-level semantic representation space for structured semantic abstraction and cross-modal alignment. Extensive experiments on the THINGS-EEG benchmark demonstrate that the proposed method achieves superior performance under subject-dependent zero-shot evaluation and improved exact retrieval under subject-independent zero-shot evaluation. Additional analyses, including layer-wise retrieval, temporal accumulation, expanded multi-image retrieval, and ablation studies, further support the effectiveness of staged decomposition and structured semantic modeling. These results suggest that explicitly modeling staged perceptual, semantic, and integrative representations provides an effective neuroscience-inspired framework for EEG-based visual decoding.

cs.CV↗

AgentMark: Utility-Preserving Behavioral Watermarking for Agents

LLM-based agents are increasingly deployed to autonomously solve complex tasks, raising urgent needs for IP protection and regulatory provenance. While content watermarking effectively attributes LLM-generated outputs, it fails to directly identify the high-level planning behaviors (e.g., tool and subgoal choices) that govern multi-step execution. Critically, watermarking at the planning-behavior layer faces unique challenges: minor distributional deviations in decision-making can compound during long-term agent operation, degrading utility, and many agents operate as black boxes that are difficult to intervene in directly. To bridge this gap, we propose AgentMark, a behavioral watermarking framework that embeds multi-bit identifiers into planning decisions while preserving utility. It operates by eliciting an explicit behavior distribution from the agent and applying distribution-preserving conditional sampling, enabling deployment under black-box APIs while remaining compatible with action-layer content watermarking. Experiments across embodied, tool-use, and social environments demonstrate practical multi-bit capacity, robust recovery from partial logs, and utility preservation. The code is available at https://github.com/Tooooa/AgentMark.

cs.CR↗

Diffusion-Guided Adversarial Perturbation Injection for Generalizable Defense Against Facial Manipulations

Recent advances in GAN and diffusion models have significantly improved the realism and controllability of facial deepfake manipulation, raising serious concerns regarding privacy, security, and identity misuse. Proactive defenses attempt to counter this threat by injecting adversarial perturbations into images before manipulation takes place. However, existing approaches remain limited in effectiveness due to suboptimal perturbation injection strategies and are typically designed under white-box assumptions, targeting only simple GAN-based attribute editing. These constraints hinder their applicability in practical real-world scenarios. In this paper, we propose AEGIS, the first diffusion-guided paradigm in which the AdvErsarial facial images are Generated for Identity Shielding. We observe that the limited defense capability of existing approaches stems from the peak-clipping constraint, where perturbations are forcibly truncated due to a fixed $L_\infty$-bounded. To overcome this limitation, instead of directly modifying pixels, AEGIS injects adversarial perturbations into the latent space along the DDIM denoising trajectory, thereby decoupling the perturbation magnitude from pixel-level constraints and allowing perturbations to adaptively amplify where most effective. The extensible design of AEGIS allows the defense to be expanded from purely white-box use to also support black-box scenarios through a gradient-estimation strategy. Extensive experiments across GAN and diffusion-based deepfake generators show that AEGIS consistently delivers strong defense effectiveness while maintaining high perceptual quality. In white-box settings, it achieves robust manipulation disruption, whereas in black-box settings, it demonstrates strong cross-model transferability.

cs.CR↗

FAST Observations of Wave-like Structures in the Radio Dynamic Spectrum of AD Leo

M-dwarf flare stars like AD Leo are laboratories for studying intense magnetic activities. The coherent radio bursts they produce are powerful probes of stellar coronal plasma and magnetic fields. In this study, we present high-resolution observations of AD Leo from the Five-hundred-meter Aperture Spherical radio Telescope (FAST) that reveal wave-like structures in its radio dynamic spectrum. The observations show trains of short-duration, narrowband sub-bursts where the central frequency, frequency drift rate, and flux density are all simultaneously modulated with a period of 1.53 s. Notably, modulation of the central frequency is approximately in-phase with that of the drift rate but roughly in anti-phase with that of the flux density. Furthermore, the amplitude of the frequency modulation grows with an e-fold timescale of 2.4 s. We interpret the observed sinusoidal frequency modulations as a possible signature of a magnetohydrodynamic (MHD) wave in the stellar corona. Our work provides a window into stellar coronal seismology and offers an opportunity to infer the local plasma environment via the MHD wave model.

astro-ph.SR↗