arXiv subjects

Tong Liu

Publications and source records attributed to Tong Liu.

At least 73 records · Page 4Linked to original sources

arXiv2026

All-domain Moveline Evolution Network for Click-Through Rate Prediction

E-commerce app users exhibit behaviors that are inherently logically consistent. A series of multi-scenario user behaviors interconnect to form the scene-level all-domain user moveline, which ultimately reveals the user's true intention. Traditional CTR prediction methods typically focus on the item-level interaction between the target item and the historically interacted items. However, the scene-level interaction between the target item and the user moveline remains underexplored. There are two challenges when modeling the interaction with preceding all-domain user moveline: (i) Heterogeneity between items and scenes: Unlike traditional user behavior sequences that utilize items as carriers, the user moveline utilizes scenes as carriers. The heterogeneity between items and scenes complicates the process of aligning interactions within a unified representation space. (ii) Temporal misalignment of linked scene-level and item-level behaviors: In the preceding user moveline with a fixed sampling length, certain critical scene-level behaviors are closely linked to subsequent item-level behaviors. However, it is impossible to establish a complete temporal alignment that clearly identifies which specific scene-level behaviors correspond to which item-level behaviors. To address these challenges and pioneer modeling user intent from the perspective of the all-domain moveline, we propose All-domain Moveline Evolution Network (AMEN). AMEN not only transfers interactions between items and scenes to homogeneous representation spaces, but also introduces a Temporal Sequential Pairwise (TSP) mechanism to understand the nuanced associations between scene-level and item-level behaviors, ensuring that the all-domain user moveline differentially influences CTR predictions for user's favored and unfavored items. Online A/B testing demonstrates that our method achieves a +11.6% increase in CTCVR.

cs.IR↗

arXiv2026

Next Interest Flow: A Generative Pre-training Paradigm for Recommender Systems by Modeling All-domain Movelines

Click-Through Rate (CTR) prediction has long been dominated by discriminative paradigms that optimize local decision boundaries within candidate-specific subspaces. However, these models often fail to capture the global joint distribution and the continuous structural evolution of user intent across all-domain movelines. While generative approaches attempt to model global transition patterns, existing methods suffer from discretization-induced information collapse by remapping nuanced e-commerce signals into discrete linguistic or categorical spaces, failing to preserve the topological fidelity of interest trajectories. To overcome these limitations, we propose a novel generative pre-training paradigm that models user intent as a continuous evolutionary trajectory on a high-dimensional latent interest manifold, termed the Next Interest Flow (NIF). We introduce kinematic constraints to govern this flow: Interest Diversity is achieved via tangent space decomposition, while Evolution Velocity ensures trajectory smoothness through geodesic regularization. To bridge the objective mismatch between generative pre-training and discriminative fine-tuning, we propose a bidirectional alignment strategy to synchronize semantic spaces. Furthermore, we develop a Temporal Sequential Pairwise (TSP) mechanism to instill temporal causality within the discriminative framework. We present the All-domain Moveline Evolution Network (AMEN), a unified framework implementing this pipeline. Extensive experiments on a 6.7-billion instance industrial dataset and online A/B tests on Taobao validate AMEN's superiority, achieving +0.87pt AUC gain and +11.6\% CTCVR lift.

cs.IR↗

arXiv2026

LoRM: Learning the Language of Rotating Machinery for Self-Supervised Condition Monitoring

We present LoRM (Language of Rotating Machinery), a self-supervised framework for multi-modal rotating-machinery signal understanding and real-time condition monitoring. LoRM is built on the idea that rotating-machinery signals can be viewed as a machine language: local signals can be tokenised into discrete symbolic units, and their future evolution can be predicted from observed multi-sensor context. Unlike conventional signal-processing methods that rely on hand-crafted transforms and features, LoRM reformulates multi-modal sensor data as a token-based sequence-prediction problem. For each data window, the observed context segment is retained in continuous form, while the future target segment of each sensing channel is quantised into a discrete token. Then, efficient knowledge transfer is achieved by partially fine-tuning a general-purpose pre-trained language model on industrial signals, avoiding the need to train a large model from scratch. Finally, condition monitoring is performed by tracking token-prediction errors as a health indicator, where increasing errors indicate degradation. In-situ tool condition monitoring (TCM) experiments demonstrate stable real-time tracking and strong cross-tool generalisation, showing that LoRM provides a practical bridge between language modelling and industrial signal analysis. The source code is publicly available at https://github.com/Q159753258/LormPHM.

cs.CL↗

arXiv2026

The Breakthrough of Sleep: A Contactless Approach for Accurate Sleep Stage Detection Using the Sleepal AI Lamp

Sleep staging is essential for the assessment of sleep quality and the diagnosis of sleep-related disorders. Conventional polysomnography (PSG), while considered the gold standard, is intrusive, labor-intensive, and unsuitable for long-term monitoring. This study evaluates the performance of the Sleepal AI Lamp, a contactless, radar-based consumer-grade sleep tracker, in comparison with gold-standard polysomnography (PSG), using a large-scale dataset comprising 1022 overnight recordings. We extract multi-scale respiratory and motion-related features from radar signals to train a frequency-augmented deep learning model. For the binary sleep-wake classification task, experimental results demonstrated that the model achieved an accuracy of 92.8% alongside a macro-averaged F1 score of 0.895. For four-stage classification (wake, light NREM (N1 + N2), deep NREM (N3), REM), the model achieved an accuracy of 78.5% with a Cohen's kappa coefficient of 0.695 in healthy individuals and maintained a stable accuracy of 77.2% with a kappa of 0.677 in a heterogeneous population including patients with varying severities of obstructive sleep apnea (OSA). These experimental results demonstrate that the sleep staging performance of the contactless Sleepal AI Lamp is in high agreement with expert-labeled PSG sleep stages. Our findings suggest that non-contact radar sensing, combined with advanced temporal modeling, can provide reliable sleep staging performance without requiring physical contact or wearable devices. Owing to its unobtrusive nature, ease of deployment, and robustness to long-term use, the contactless Sleepal AI Lamp shows strong potential for clinical screening, home-based sleep assessment, and continuous longitudinal sleep monitoring in real-world medical and healthcare applications.

eess.SP↗

arXiv2026

An Intertwined Short and Long GRB with 4-minute Separation

Gamma-ray bursts (GRBs), the most energetic transients in the Universe, are traditionally classified into long-duration ($T_{90}>2$ s) and short-duration ($T_{90}<2$ s) events, associated with the core collapse of massive stars (Type II) and the merger of compact binary systems (Type I), respectively. The two classes exhibit distinct observational properties that serve as key diagnostic criteria for classification. Here we report GRB 160425A, a peculiar event comprising two sub-bursts separated by four minutes: a short-duration burst ($G_1$) and a long-duration burst ($G_2$). Nearly all standard prompt-emission diagnostics, including pulse morphology, duration, hardness ratio, minimum variability timescale, spectral properties, and established empirical correlations, consistently categorize $G_1$ as a short-like (Type I, merger-origin) and $G_2$ as a long-like (Type II, collapsar-origin) GRB. The coexistence of merger and collapsar signatures in a single event challenges existing progenitor frameworks and calls for a re-evaluation of GRB classification schemes and progenitor scenarios.

astro-ph.HE↗

arXiv2026

Inherent self-consistency of the electron fraction between neutrino-dominated accretion flows and their progenitors

Stellar-mass black holes (BHs) surrounded by neutrino-dominated accretion flows (NDAFs) are a leading central engine of gamma-ray bursts (GRBs). In this work, we investigate the electron fraction distribution in NDAFs with or without disk outflows for different accretion rates, BH spins, and outflow rates. As the results, for the cases of the massive disks at relatively low accretion rates, the outer boundary of the disks are predominantly advection-cooled, yielding electron fractions of $Y_{\rm e} \sim 0.5$, as expected for massive collapsar progenitors. By contrast, in the cases of lower-mass disk at high accretion rates, neutrino cooling becomes highly efficient and mildly electron-degenerate disks emerge, characterized by $Y_{\rm e} \lesssim 0.38$ at the outer boundary of the disk, even for the strong outflows, which is consistent with materials from compact object merger scenarios. Moreover, we find that these trends remain robust across different BH spins. Consequently, the self-consistent agreement between the electron fraction properties at the outer boundaries of NDAFs and those expected from GRB progenitors provides effectively support for NDAFs serving as the GRB central engines.

astro-ph.HE↗

arXiv2026

Parameter trajectory engineering for state transfer and quantum sensing in non-Hermitian two-level systems

Exceptional points (EPs) in non-Hermitian systems give rise to enhanced sensitivity and chiral state transfer, which are important for quantum technologies. Although parameter trajectories encircling EPs can control symmetric and chiral state transfer, their robustness against practical perturbations and their role in quantum sensing remain largely unexplored. Here, we study three time-modulated parameter loops in a non-Hermitian two-level system to show how trajectory design governs state-transfer symmetry, robustness, and sensing performance. Trajectories avoiding the EP support robust symmetric transfer, while those encircling the EP yield chiral transfer governed by the topological winding number, whose robustness depends on the distance to the EP and the encircling direction. For quantum sensing, trajectory engineering enables tuning of sensitivity amplitude, time window, and parameter selectivity in both eigenvalue-based and eigenstate-based sensors. Notably, eigenstate-based sensing achieves full parameter selectivity that is unattainable with eigenvalue-based methods. Our results establish a quantitative connection between trajectory topology and system dynamics, providing a unified framework for robust state-transfer protocols and high-performance quantum sensors.

quant-ph↗

arXiv2026

$\mathrm{Cs_3V_9Te_{13}}$: A New Vanadium-Based Material with a Reuleaux-Triangle-Like Lattice and a Possible Phase Transition near 48 K

Exploring and synthesizing materials with new crystal structures provides an important route to discovering exotic quantum phenomena. However, materials with unconventional lattice geometries remain largely unexplored. Here, we report the discovery of a new vanadium-based material, $\mathrm{Cs_3V_9Te_{13}}$, featuring a Reuleaux-triangle-like lattice. Electrical transport and magnetic measurements consistently reveal an anomaly near 48 K, and this feature shows little sensitivity to the applied magnetic field. A corresponding anomaly is also observed in the Hall coefficient near 48 K, indicating a marked change in the carrier response. In addition, temperature-dependent x-ray diffraction results indicate no obvious structural change across 48 K. Taken together, these results suggest that the anomaly is not induced by the structural transition, but associated to a possible electronic and/or magnetic phase transition. High-pressure transport measurements and first-principles calculations further reveal a highly tunable electronic state in $\mathrm{Cs_3V_9Te_{13}}$, with the kagome-like electronic feature and pressure-suppressed antiferromagnetism. These results demonstrate this material, with its structurally novel Reuleaux-triangle-like lattice, as a new platform for exploring the interplay between nontrivial lattice geometry and emergent physical phenomena.

cond-mat.mtrl-sci↗

arXiv2026

Torsion graded pieces of Nyggard filtration for crystalline representation

Let $K$ be a unramified $p$-adic field with the absolute Galois group $G_K$ and $T$ a crystalline $\mathbb Z_p$-representation of $G_K$. We study the graded pieces of integral filtration on $D_{\rm dR}(T)$ given by Nyggard filtration of the attached Breuil-Kisin module of $T$. We show that the $i$-graded piece has nontrivial $p$-torsion only if $ i = r_j +m p$ for a Hodge-Tate weight $ r_j$ of $T$ and $m$ a positive integer.

math.NT↗

arXiv2026

Disentangling Tensor Network States with Deep Neural Network

We introduce Neural Tensor Network States ($ν$TNS), a variational many-body wave-function ansatz that integrates deep neural networks with tensor-network architectures. In the $ν$TNS framework, a neural network serves as a disentangler of the wave-function, transforming the physical degrees of freedom into renormalized variables with much less entanglement. The renormalized state is then efficiently encoded by a back-flow tensor network. This construction yields a compact yet highly expressive representation of strongly correlated quantum states. Using convolutional neural networks combined with matrix product states as a concrete implementation, we obtain state-of-the-art variational energies for the spin-$1/2$ $J_1$-$J_2$ Heisenberg model on the square lattice at the highly frustrated point $J_2/J_1=0.5$, for systems up to $20\times 20$ with periodic boundary conditions. Finite-size scaling of spin, dimer, and plaquette correlations exhibits power-law decay without magnetic or valence-bond long-range order, consistent with a gapless quantum spin-liquid ground state at that point.This $ν$TNS framework is flexible and naturally extensible to other neural and tensor-network structures, offering a general platform for investigating strongly correlated quantum many-body systems.

cond-mat.str-el↗

arXiv2026

Multi-instrument constraints on a hemispherically asymmetric positive ionospheric storm in the 60-180 deg E sector during the 12-13 November 2025 geomagnetic storm

Geomagnetic storms drive complex ionospheric responses through coupled electrodynamic and thermospheric processes, yet attributing storm-time TEC perturbations to specific mechanisms remains challenging. We investigate the ionospheric response to the 12-13 November 2025 intense geomagnetic storm (Dst minimum = -214 nT) in the 60-180 deg E sector using a coordinated multi-instrument dataset comprising JPL GIM TEC, dense regional GNSS networks, continuous BeiDou GEO links, COSMIC-2 radio occultation, ground ionosondes, Swarm in-situ electron density, HF Doppler soundings, and TIMED/GUVI thermospheric composition observations. The observations reveal a dayside-dominant positive TEC storm with pronounced hemispheric asymmetry, where Northern Hemisphere mid-to-low latitudes exhibit stronger and longer-lasting enhancement than the Southern Hemisphere. Joint analysis of radio occultation, ionosonde, and Swarm data indicates that the enhancement is density-dominated with NmF2 and foF2 increases but with no coherent, sector-scale peak-height uplift in hmF2 or h'F2, posing challenges for uplift-only electrodynamic interpretations. Coherent large-scale traveling ionospheric disturbances propagate across the equator during UT 1-6, while HF Doppler oscillations maximize later during UT 6-24, revealing a timing offset between integrated TEC responses and reflection-height dynamics. Southern Hemisphere O/N2 ratio depletion observed by TIMED/GUVI provides compositional context consistent with the faster positive-phase decay there, although concurrent Northern Hemisphere GUVI coverage is limited during this interval. These findings highlight the value of multi-observable diagnostics for developing testable constraints on storm-time mechanisms and improving sector-specific space weather nowcasting capabilities.

physics.geo-ph↗

arXiv2026

Scalable Training of Mixture-of-Experts Models with Megatron Core

Scaling Mixture-of-Experts (MoE) training introduces systems challenges absent in dense models. Because each token activates only a subset of experts, this sparsity allows total parameters to grow much faster than per-token computation, creating coupled constraints across memory, communication, and computation. Optimizing one dimension often shifts pressure to another, demanding co-design across the full system stack. We address these challenges for MoE training through integrated optimizations spanning memory (fine-grained recomputation, offloading, etc.), communication (optimized dispatchers, overlapping, etc.), and computation (Grouped GEMM, fusions, CUDA Graphs, etc.). The framework also provides Parallel Folding for flexible multi-dimensional parallelism, low-precision training support for FP8 and NVFP4, and efficient long-context training. On NVIDIA GB300 and GB200, it achieves 1,233/1,048 TFLOPS/GPU for DeepSeek-V3-685B and 974/919 TFLOPS/GPU for Qwen3-235B. As a performant, scalable, and production-ready open-source solution, it has been used across academia and industry for training MoE models ranging from billions to trillions of parameters on clusters scaling up to thousands of GPUs. This report explains how these techniques work, their trade-offs, and their interactions at the systems level, providing practical guidance for scaling MoE models with Megatron Core.

cs.DC↗

arXiv2026

Log prismatic $F$-crystals and purity

Our goal is to study $p$-adic local systems on a rigid-analytic variety with semistable formal model. We prove that such a local system is semistable if and only if so are its restrictions to the points corresponding to the irreducible components of the special fiber. For this, the main body of the paper concerns analytic prismatic $F$-crystals on the absolute logarithmic prismatic site of a semistable $p$-adic log formal scheme. Analyzing Breuil-Kisin log prisms, we obtain a prismatic purity theorem and deduce the above purity theorem for semistable local systems.

math.NT↗

arXiv2026

Dual-Space Invariance as a Universal Criterion for Multifractal Critical States

In Anderson localization, eigenstates of disordered quantum systems are broadly classified as extended, localized, or critical. Although critical states exhibit multifractal character, a precise and operational criterion for their identification remains an open challenge, as Lyapunov exponents in real space cannot uniquely distinguish them from extended states. Here we address this challenge by asserting that critical states are uniquely characterized by an emergent dual-space invariance between position and momentum space. Building on the Liu--Xia criterion of the simultaneous vanishing of Lyapunov exponents ($γ=γ_m=0$), we show that this dual-space invariance extends beyond Lyapunov exponents and governs wavefunction scaling, revealing a fundamental property inaccessible from either space alone. Through numerical simulations, we demonstrate that the inverse participation ratio exhibits matching scaling behavior in position and momentum space for critical states, in sharp contrast to extended and localized states, which display a pronounced asymmetry between the two spaces. This dual-space invariance provides a direct, robust, and universal criterion for identifying multifractal critical states. Our results establish a fundamental principle of Anderson criticality and open new avenues for its detection in modern quantum simulation platforms.

cond-mat.dis-nn↗

arXiv2026

MoE Parallel Folding: Heterogeneous Parallelism Mappings for Efficient Large-Scale MoE Model Training with Megatron Core

Mixture of Experts (MoE) models enhance neural network scalability by dynamically selecting relevant experts per input token, enabling larger model sizes while maintaining manageable computation costs. However, efficient training of large-scale MoE models across thousands of GPUs presents significant challenges due to limitations in existing parallelism strategies. We introduce an end-to-end training framework for large-scale MoE models that utilizes five-dimensional hybrid parallelism: Tensor Parallelism, Expert Parallelism, Context Parallelism, Data Parallelism, and Pipeline Parallelism. Central to our approach is MoE Parallel Folding, a novel strategy that decouples the parallelization of attention and MoE layers in Transformer models, allowing each layer type to adopt optimal parallel configurations. Additionally, we develop a flexible token-level dispatcher that supports both token-dropping and token-dropless MoE training across all five dimensions of parallelism. This dispatcher accommodates dynamic tensor shapes and coordinates different parallelism schemes for Attention and MoE layers, facilitating complex parallelism implementations. Our experiments demonstrate significant improvements in training efficiency and scalability. We achieve up to 49.3% Model Flops Utilization (MFU) for the Mixtral 8x22B model and 39.0% MFU for the Qwen2-57B-A14B model on H100 GPUs, outperforming existing methods. The framework scales efficiently up to 1,024 GPUs and maintains high performance with sequence lengths up to 128K tokens, validating its effectiveness for large-scale MoE model training. The code is available in Megatron-Core.

cs.LG↗

arXiv2026

Imprints of gravitational waves from magnetar spindown in GRB X-ray afterglows

Given that newborn magnetars are considered potential central engines of gamma-ray bursts (GRBs), there is strong motivation to identify gravitational wave (GW) signatures within GRB samples. If the X-ray afterglow of a GRB is powered by a magnetar, and the initial spindown of the magnetar is dominated by the GW radiation induced by $r$-mode instability or magnetic-field-induced deformation, the decay of the X-ray flux would record the information of the GW radiation. We find that GRB 130603B potentially represents a rare and precious case where the spindown of the central magnetar is dominated in-turn by $r$-mode and magnetic distortion-induced GW radiation. By fitting the X-ray light curve of GRB 130603B in this model, we obtain the initial spin period of magnetar $\sim 5.3\times 10^{-4}$ s, the effective dipole magnetic field strength $\sim 5.2\times 10^{14}$ G, the ellipticity of the magnetar $\sim 1.3\times 10^{-4}$, and the amplitude of $r$-mode oscillation $\sim3.3\times 10^{-2}$. It may serve as a reliable approach for investigating neutron star physics by comparing the parameters estimated using the method presented in this manuscript with those obtained from future GW observations.

astro-ph.HE↗

arXiv2026

Exploring the central engines of gamma-ray bursts from prompt light curves

Hyperaccreting stellar-mass black hole systems are leading candidates for the central engines of gamma-ray bursts (GRBs). Their jets are thought to be powered by either the Blandford-Znajek (BZ) process or neutrino-dominated accretion flows (NDAFs), but discriminating between these mechanisms remains challenging. To address this, we propose using the luminosity decay slope (parameter d) of GRB light curves to distinguish between the BZ and NDAF mechanisms, thereby linking the light-curve morphology to the central engine physics. By analysing 85 single-peaked GRBs with fast-rise, exponential-decay (FRED) profiles observed by Swift/BAT using 64 ms background-subtracted light curves, we fit the decay slope (parameter d) with the empirical Kocevski-Ryde-Liang (KRL) function and compare the results with theoretical predictions for the BZ (d approximately 1.67) and the NDAF (d approximately 3.7 to 7.8) mechanisms. We find that the decay slope (parameter d) can differentiate central engine mechanisms, with 15 GRBs consistent with the BZ mechanism and 22 supporting the NDAF mechanism. However, most events exhibit slopes within the range between 2 and 4, suggesting a hybrid of mechanisms, with NDAF being dominant.

astro-ph.HE↗

arXiv2026

Multi-Agent Learning Path Planning via LLMs

The integration of large language models (LLMs) into intelligent tutoring systems offers transformative potential for personalized learning in higher education. However, most existing learning path planning approaches lack transparency, adaptability, and learner-centered explainability. To address these challenges, this study proposes a novel Multi-Agent Learning Path Planning (MALPP) framework that leverages a role- and rule-based collaboration mechanism among intelligent agents, each powered by LLMs. The framework includes three task-specific agents: a learner analytics agent, a path planning agent, and a reflection agent. These agents collaborate via structured prompts and predefined rules to analyze learning profiles, generate tailored learning paths, and iteratively refine them with interpretable feedback. Grounded in Cognitive Load Theory and Zone of Proximal Development, the system ensures that recommended paths are cognitively aligned and pedagogically meaningful. Experiments conducted on the MOOCCubeX dataset using seven LLMs show that MALPP significantly outperforms baseline models in path quality, knowledge sequence consistency, and cognitive load alignment. Ablation studies further validate the effectiveness of the collaborative mechanism and theoretical constraints. This research contributes to the development of trustworthy, explainable AI in education and demonstrates a scalable approach to learner-centered adaptive instruction powered by LLMs.

cs.AI↗