arXiv subjects

Zhihao Xu

Publications and source records attributed to Zhihao Xu.

At least 19 recordsLinked to original sources

arXiv2026

Even--odd classification of reentrant topology in a Su--Schrieffer--Heeger chain with integer-power quasiperiodic hopping

We investigate reentrant topology in a one-dimensional Su--Schrieffer--Heeger chain with integer-power quasiperiodically modulated intracell hopping. The modulation is controlled by a positive integer exponent $n$ and a tunable parameter $β$, which interpolates between a smooth integer-power quasiperiodic profile and a sign-function limit. Combining the zero-mode inverse-localization-length criterion with a real-space topological indicator, we determine the phase diagrams in the $β\to0$, $β\to\infty$, and finite-$β$ regimes. We find an even--odd classification of reentrant topological windows governed by the dc component and support of the full modulation profile. For positive modulation strength, odd powers yield a zero-mean sign-changing profile and can induce reentrance from the clean trivial regime $|t_1|>1$, whereas even powers yield a positive-mean non-negative profile and allow reentrance only from the negative clean trivial regime $t_1<-1$. Although even-power profiles contain sign-changing fluctuations after subtracting their positive average, the full modulation profile remains non-negative. This even--odd structure of the full modulation profile provides a control knob for topological Anderson-like reentrant phases. We further discuss the associated bulk localization properties and show that the phase diagrams are robust against moderate hopping fluctuations, suggesting a feasible realization in electrical circuits.

cond-mat.dis-nn↗

arXiv2026

Learning to Detect Unseen Jailbreak Attacks in Large Vision-Language Models

Despite extensive alignment efforts, Large Vision-Language Models (LVLMs) remain vulnerable to jailbreak attacks. To mitigate these risks, existing detection methods are essential, yet they face two major challenges: generalization and accuracy. While learning-based methods trained on specific attacks fail to generalize to unseen attacks, learning-free methods based on hand-crafted heuristics suffer from limited accuracy and reduced efficiency. To address these limitations, we propose Learning to Detect (LoD), a learnable framework that eliminates the need for any attack data or hand-crafted heuristics. LoD operates by first extracting layer-wise safety representations directly from the model's internal activations using Multi-modal Safety Concept Activation Vectors classifiers, and then converting the high-dimensional representations into a one-dimensional anomaly score for detection via a Safety Pattern Auto-Encoder. Extensive experiments demonstrate that LoD consistently achieves state-of-the-art detection performance (AUROC) across diverse unseen jailbreak attacks on multiple LVLMs, while also significantly improving efficiency. Code is available at https://github.com/ShuangLiangX/Learning-to-Detect.

cs.CR↗

arXiv2026

Limitations of SVD-Based Diagnostics for Non-Hermitian Many-Body Localization with Time-Reversal Symmetry

Singular value decomposition (SVD) provides a convenient way to construct Hermitian-like diagnostics for non-Hermitian many-body systems, but its reliability for locating many-body localization (MBL) transitions remains unclear, particularly in systems preserving time-reversal symmetry (TRS). We benchmark SVD-based diagnostics against exact diagonalization (ED) in TRS-preserving non-Hermitian hard-core-boson chains with nonreciprocal hopping, considering quasiperiodic, random-disorder, and Stark potentials. We compare level statistics, half-chain entanglement entropy, inverse participation ratio, and spectral form factors. For the quasiperiodic and random-disorder models, ED-based entanglement and IPR yield mutually consistent finite-size transition estimates, whereas the corresponding SVD-based estimates are systematically shifted to larger disorder strengths and can lead to different phase assignments. The discrepancy is also reflected in the spectral form factors, where the ED-based dissipative spectral form factor and the SVD-based singular form factor can indicate different spectral regimes at the same parameters. In contrast, for the clean Stark model, ED and SVD give consistent transition estimates. We identify the origin of this model dependence as the fact that SVD probes the auxiliary Hermitian operator $\hat H^\dagger\hat H$, rather than the intrinsic right-eigenstate structure of $\hat H$; consequently, SVD can be quantitatively reliable only when the corresponding bulk-state structures remain aligned. Our results show that SVD-based diagnostics can capture qualitative RMT-to-Poisson trends, but are not generically reliable quantitative probes of MBL transitions in TRS-preserving non-Hermitian many-body systems.

cond-mat.dis-nn↗

arXiv2026

Gleam: Adaptive Network-Efficient CUDA API Remoting for Cross-Device GPU Sharing over LANs

This paper aims to enable computation- and communication-efficient GPU sharing across devices within local area networks (LANs), facilitating ubiquitous AI inference on heterogeneous personal devices. We achieve distributed task offloading via CUDA API remoting. However, beyond raw computation, network constraints emerge as the primary bottleneck: limited bandwidth, high-frequency API invocations, and cross-task contention significantly hinder performance. To address these challenges, we propose Gleam, a novel and network-efficient framework for task-generic GPU sharing across local-area CUDA devices, with three key contributions. First, we reduce bandwidth overhead in CUDA API remoting through automatic model weight caching, and mitigate accumulated latency from frequent API calls by asynchronous execution. Second, we design a runtime task scheduler that dynamically determines API remoting pairs between LAN clients and servers, explicitly accounting for both network conditions and GPU resource contention under parallel workloads. Finally, we introduce dedicated mechanisms to ensure CUDA context consistency across distributed executions. Extensive experiments on heterogeneous NVIDIA GPUs and diverse AI workloads show Gleam consistently outperforms state-of-the-art baselines, achieving 1.4-24.2 times improvements in API remoting efficiency and up to 1.79 times higher system throughput.

cs.DC↗

arXiv2026

Distributed Quantum Approximate Optimization Algorithm on a Quantum-Centric Supercomputing Architecture

Quantum approximate optimization algorithm (QAOA) has shown promise in solving combinatorial optimization problems by providing quantum speedup on near-term gate-based quantum computing systems. However, QAOA faces challenges for high-dimensional problems due to the large number of qubits required and the complexity of deep circuits, limiting its scalability for real-world applications. In this study, we present a distributed QAOA (DQAOA), which leverages distributed computing strategies to decompose a large computational workload into smaller tasks that require fewer qubits and shallower circuits than necessitated to solve the original problem. These sub-problems are processed using a combination of high-performance and quantum computing resources. The global solution is iteratively updated by aggregating sub-solutions, allowing convergence toward the optimal solution. We demonstrate that DQAOA can handle considerably large-scale optimization problems (e.g., 1,000-bit problem) achieving a high solution quality and short time-to-solution ($\sim$276 s), outperforming existing strategies. Furthermore, we realize DQAOA on a quantum-centric supercomputing architecture, paving the way for practical applications of gate-based quantum computers in real-world optimization tasks. To extend DQAOA's applicability to materials science, we further develop an active learning algorithm integrated with our DQAOA (AL-DQAOA), which involves machine learning, DQAOA, and active data production in an iterative loop. We successfully optimize photonic structures using AL-DQAOA, indicating that solving real-world optimization problems using gate-based quantum computing is feasible. We expect the proposed DQAOA to be applicable to a wide range of optimization problems and AL-DQAOA to find broader applications in material design.

cs.DC↗

arXiv2026

Deep-Ocean Application-Specific Neutrino Experiment

This report introduces the concept, prototype design, projected costs, and scientific goals of a mobile experiment for detecting geoneutrinos originating from uranium and thorium decay chains in the Earth's mantle. This will constrain the planet's radiogenic heat production and unearth its geochemical makeup. This design of a deep-ocean mobile neutrino experiment, which is not mirrored by any active or planned experiments, supports physics and geoscience's goal of multi-modal data on the Earth's internal composition and structure. Based on geoscientific studies, this design is expected to achieve a 50--100-fold reduction in crustal background compared to similarly sized continental detectors, thereby enabling direct measurements of mantle geoneutrinos. The multiple stereoscopic projections enabled by the detector's unique mobility can map spatial variations in heat-producing elements within the mantle. Beyond discussing the design, we report on our collaboration's most recent hardware developments in the active prototyping of this detector. We briefly highlight the potential multiuse and interdisciplinary nature of this detector.

physics.ins-det↗

arXiv2026

Mapping deep-mantle compositional heterogeneity using a directional geoneutrino detector

Determining the spatial distribution of heat-producing elements (HPEs) within the Earth is critical for understanding the planet's thermal and chemical evolution. A central debate is whether the deep mantle, particularly the Large Low-Velocity Provinces (LLVPs), retains anomalous, radiogenically enriched reservoirs. While mapping surface variations in geoneutrino flux offers a direct probe of Earth's internal radioactivity, current continental-located detectors measure only the angle-integrated flux. This limitation creates a fundamental parameter degeneracy, rendering it impossible to distinguish a chemically homogeneous mantle from a heterogeneous one. In this study, we quantify the potential of directional geoneutrino detection to overcome this limitation. By evaluating realistic LLVP geometries under the experimental framework of the proposed Ocean Bottom Detector (OBD), we demonstrate that resolving the incoming direction of geoneutrinos can successfully break the non-uniqueness inherent in rate-only measurements. These results indicate that future directional geoneutrino measurements could help determine whether LLVPs host enhanced HPE abundances and assess their contribution to Earth's radiogenic heat budget. Such measurements would provide a new observational constraint on the chemical heterogeneity of the deep mantle and its role in Earth's long-term thermal evolution.

physics.geo-ph↗

arXiv2026

Riesz--Fejér type Inequalities for $α$-Harmonic Functions in the Unit Ball

In this paper, we establish Riesz--Fejér type inequalities for the $α$-harmonic functions $f=P_α[f^*]$ in $\mathbb B^n$, where $f^*\in L^{p}(\mathbb{S}^{n-1})$ and $1 -1$, we prove the existence of a constant $\mathcal{C}_{n,p,α}$ such that $\int_{-1}^{1} |f(rη)|^p(1-r^2)^{n-2}\,dr \leq \mathcal{C}_{n,p,α} \int_{\mathbb S^{n-1}}|f^*(ξ)|^p\,dσ(ξ)$. Moreover, in the range $α>\max\left\{ -\frac{n-1}{p},\,n-2-\frac{2(n-1)}{p} \right\}$, we determine the sharp constant explicitly. The result generalize and extend the corresponding results of Ahmed et al. (J. Math. Anal. Appl., 563:13, 2026), Hu et al. (Anal. Math. 51:15, 2025) and Long (arXiv: 2410.12137).

math.CV↗

arXiv2026

Generalized Aubry-André-Harper model with power-law quasiperiodic potentials

We investigate a generalized Aubry-André-Harper (AAH) model with non-reciprocal hopping and power-law quasiperiodic potentials $V(i) = V\left[ \cos(2πβi) \right]^p$. Our study reveals that the interplay between nonreciprocity, quasiperiodicity, and the power-law exponent $p$ gives rise to a variety of phase transitions and localization phenomena. In the Hermitian case, the system undergoes a direct transition from extended to localized phases for $p=1, 2$, while for $p \geq 3$, an intermediate mixed phase emerges, characterized by the coexistence of extended and localized states and the presence of mobility edges. Importantly, we find that prominent high-IPR states associated with well-resolved spectral gaps appear at specific energy levels, whose positions are captured by the relation $x_n = nβ- \lfloor nβ\rfloor$, for low-order $n$. In the non-Hermitian regime, the energy spectrum becomes complex and the $\mathcal{PT}$ transition coincides with the extended-to-localized phase boundary for $p = 1, 2$, whereas for $p \geq 3$, $\mathcal{PT}$-symmetry breaking occurs at the mixed-to-localized phase transition. This work reveals how power-law quasiperiodic potentials and non-reciprocal hopping govern phase transitions, providing new insight into localization phenomena of quasiperiodic systems.

cond-mat.dis-nn↗

arXiv2026

Towards imaging Earth's large-scale structures by directional geoneutrino detection with Ocean Bottom Detector

Geoneutrinos, electron antineutrinos produced by radioactive decays of heat-producing elements (HPEs) within the Earth, provide unique insights into Earth's interior and heat budget since their first detection in 2005 by KamLAND. Conventional geoneutrino detectors currently provide integrated global information and lack the capability to spatially resolve structures deep within the Earth. Here, we evaluate the ability of angular-sensitive geoneutrino detectors to distinguish between homogeneous and heterogeneous mantle models, focusing on Large Low Shear Velocity Provinces (LLSVPs). Our results show that LLSVPs enriched in Th and U yield a distinct flux of geoneutrinos with distinctive angular patterns. An oceanic site above the Pacific LLSVP is considered a particularly favorable detector location. The Ocean Bottom Detector (OBD) project aims to leverage this spatial resolving advantage by deploying a kiloton-scale liquid scintillator detector directly on the ocean floor, enabling unprecedented sensitivity for mantle geoneutrino detection. These findings demonstrate the critical role of combining geophysical and geochemical data to guide detector site selection, ultimately improving constraints on Earth's internal heat and the HPE distribution.

physics.geo-ph↗

arXiv2026

Information-Theoretic Decomposition for Multimodal Interaction Learning

Multimodal learning hinges on capturing redundant, unique, and synergistic information across modalities, which collectively constitute multimodal interactions. A critical yet underexplored challenge is that these implicit interactions vary dynamically across samples. In this work, we present the first systematic, information-theoretic analysis highlighting why learning these dynamic, sample-specific interactions is critical for effective multimodal learning. Our analysis further reveals deficits in conventional paradigms at learning these distinct interaction types: modality ensemble approaches struggle to capture synergy, while joint learning paradigms often under-utilize redundant information. This highlights the need for an approach that can adaptively learn from different interaction types on a per-sample basis. To this end, we propose Decomposition-based Multimodal Interaction Learning (DMIL), a novel paradigm that explicitly models and learns from sample-specific interactions. First, we design a variational decomposition architecture to isolate the constituent interaction components. Second, we employ a new learning strategy that leverages these explicit interaction components in a fine-tuning process to achieve comprehensive interaction learning. Extensive experiments across diverse tasks and architectures demonstrate that DMIL consistently achieves superior performance by adapting to holistic sample-specific interactions. Our framework is flexible and broadly applicable, establishing an interaction-centric paradigm for multimodal learning. The code is available at https://github.com/GeWu-Lab/DMIL.

cs.LG↗

arXiv2026

Multiple reentrant topological windows induced by generalized Bernoulli disorder

We investigate reentrant topological transitions in a one-dimensional Su-Schrieffer-Heeger chain with generalized Bernoulli disorder in the intradimer hopping amplitudes. Owing to its independently tunable values and probabilities, the multivalued disorder distribution provides a direct way to control the topological phase diagram. We show that increasing the disorder strength can split the nontrivial regime into multiple disconnected topological windows, whose number, widths, and locations are determined by the distribution parameters. The phase boundaries are derived analytically from the zero-mode inverse localization length and are governed by a weighted geometric mean of the disordered hopping amplitudes, in agreement with numerical results from the reflection-matrix topological quantum number and the real-space winding number. We also show that the mean chiral displacement dynamically identifies these reentrant windows. These results demonstrate how multivalued random disorder can organize and tune reentrant topological behavior in one-dimensional chiral lattices.

physics.optics↗

arXiv2026

Impurity-induced loss bursts from anomalous scale-free localization in a non-Hermitian dissipative lattice

We identify anomalous scale-free localization and the associated impurity-induced loss bursts in a non-Hermitian dissipative cross-stitch lattice. By a local basis rotation, the model is mapped onto an effective non-Hermitian Su-Schrieffer-Heeger lattice, where local impurities act as tunable effective boundaries. For the parameter choice considered here, tuning the impurity strength $η$ connects two effective open-boundary-condition-like limits, reached for $η\to0$ and $η\to\infty$, through generalized-boundary-condition regimes and the impurity-free periodic-boundary-condition point at $η=1$. For finite $η\notin\{0,1\}$, the spectral loops remain separated from the real-energy axis, while the eigenstates exhibit scale-free localization pinned by the impurity. Unlike conventional impurity-induced scale-free localization, the Lyapunov exponent depends explicitly on the eigenenergy, making the localization strength eigenstate dependent. We further show that this anomalous eigenmode structure produces an impurity-induced loss burst: the long-time integrated dissipation probability is strongly enhanced near an impurity-generated effective boundary even when the initial wave packet is far away. In the single-impurity case, the burst region consists of the impurity site and its adjacent effective-boundary site, and the effect occurs without imaginary-gap closing. For multiple impurities, local burst regions emerge around all impurities, while the dominant burst boundary is selected by the initial wave-packet position and the nonreciprocal drift direction. These results connect anomalous scale-free localization with controllable dissipation dynamics in non-Hermitian lattices.

quant-ph↗

arXiv2026

Exceptional-point-constrained locking of boundary-sensitive topological transitions in chiral non-Hermitian SSH-type lattices

Topological transitions in non-Hermitian systems are generally boundary sensitive: a point-gap winding transition under periodic boundary condition (PBC) and a non-Bloch bulk real-line-gap transition under open boundary condition (OBC) at $\mathrm{Re}(E)=0$ are governed by different spectra and therefore need not coincide. Here we show, for a class of chiral non-Hermitian Su--Schrieffer--Heeger (SSH)-type lattices, that these two criticalities can be locked by an exceptional-point-constrained (EP-constrained) parameter evolution. The key requirement is not the occurrence of isolated exceptional points, but the persistence of a zero-energy Bloch degeneracy along the entire sweep, which is generically exceptional in the non-Hermitian regime. In an analytically tractable limit of an extended non-Hermitian SSH chain, the EP-constrained manifolds and both transition boundaries are obtained in closed form, making the locking explicit. Away from this limit, numerical generalized-Brillouin-zone (GBZ) calculations confirm the correspondence for representative constrained sweeps, whereas unconstrained paths show that isolated exceptional points or Hermitian degeneracies do not enforce locking. We further verify the mechanism in a spinful four-band extension with branch-resolved GBZs, including strongly branch-imbalanced regimes. These results establish a path-dependent diagnostic principle: along EP-constrained sweeps in this SSH-type class, changes in PBC point-gap winding can indicate OBC non-Bloch bulk real-line-gap transitions and the corresponding changes in zero-energy boundary modes.

physics.optics↗

arXiv2026

Think-with-Rubrics: From External Evaluator to Internal Reasoning Guidance

Rubrics have been extensively utilized for evaluating unverifiable, open-ended tasks, with recent research incorporating them into reward systems for reinforcement learning. However, existing frameworks typically treat rubrics only as external evaluator disjointed from the policy's primary reasoning trace. Such design confines rubrics to post-hoc measurement, leaving them unable to actively guide the model's generation process. In this work, we introduce Think-with-Rubrics, a novel paradigm for instruction following tasks. Think-with-Rubrics integrates rubric generation into the reasoning context, transforming the rubric from an independent artifact into an internal guidance of LLM's generation. During training, LLM sequentially generates a rubric followed by a response, while a trained rubric verifier provides joint supervision by evaluating the consistency between the answer and the self-generated / golden rubrics. Experiments across multiple benchmarks demonstrate that Think-with-Rubrics consistently outperforms the Rubric-as-Reward baseline supervised by golden rubrics by an average of 3.87 points. We have also discussed the mechanism by which Think-with-Rubrics enhances model performance. Experimental results demonstrate that supervision from golden rubrics and self-generated rubrics enhances the performance of Think-with-Rubrics by improving the quality of self-generated rubrics and increasing the internal consistency of responses respectively.

cs.CL↗

arXiv2026

DriVerse: Navigation World Model for Driving Simulation via Multimodal Trajectory Prompting and Motion Alignment

This paper presents DriVerse, a generative model for simulating navigation-driven driving scenes from a single image and a future trajectory. Previous autonomous driving world models either directly feed the trajectory or discrete control signals into the generation pipeline, leading to poor alignment between the control inputs and the implicit features of the 2D base generative model, which results in low-fidelity video outputs. Some methods use coarse textual commands or discrete vehicle control signals, which lack the precision to guide fine-grained, trajectory-specific video generation, making them unsuitable for evaluating actual autonomous driving algorithms. DriVerse introduces explicit trajectory guidance in two complementary forms: it tokenizes trajectories into textual prompts using a predefined trend vocabulary for seamless language integration, and converts 3D trajectories into 2D spatial motion priors to enhance control over static content within the driving scene. To better handle dynamic objects, we further introduce a lightweight motion alignment module, which focuses on the inter-frame consistency of dynamic pixels, significantly enhancing the temporal coherence of moving elements over long sequences. With minimal training and no need for additional data, DriVerse outperforms specialized models on future video generation tasks across both the nuScenes and Waymo datasets. The code and models will be released to the public.

cs.RO↗

arXiv2026

U-ViLAR: Uncertainty-Aware Visual Localization for Autonomous Driving via Differentiable Association and Registration

Accurate localization using visual information is a critical yet challenging task, especially in urban environments where nearby buildings and construction sites significantly degrade GNSS (Global Navigation Satellite System) signal quality. This issue underscores the importance of visual localization techniques in scenarios where GNSS signals are unreliable. This paper proposes U-ViLAR, a novel uncertainty-aware visual localization framework designed to address these challenges while enabling adaptive localization using high-definition (HD) maps or navigation maps. Specifically, our method first extracts features from the input visual data and maps them into Bird's-Eye-View (BEV) space to enhance spatial consistency with the map input. Subsequently, we introduce: a) Perceptual Uncertainty-guided Association, which mitigates errors caused by perception uncertainty, and b) Localization Uncertainty-guided Registration, which reduces errors introduced by localization uncertainty. By effectively balancing the coarse-grained large-scale localization capability of association with the fine-grained precise localization capability of registration, our approach achieves robust and accurate localization. Experimental results demonstrate that our method achieves state-of-the-art performance across multiple localization tasks. Furthermore, our model has undergone rigorous testing on large-scale autonomous driving fleets and has demonstrated stable performance in various challenging urban scenarios.

cs.CV↗

arXiv2026

Anomalous Localization and Mobility Edges in Non-Hermitian Quasicrystals with Disordered Imaginary Gauge Fields

Localization in non-Hermitian quasicrystals can differ fundamentally from its Hermitian counterpart when non-reciprocity is spatially disordered. Here we study a one-dimensional non-Hermitian Aubry-André-Harper chain with a Bernoulli imaginary gauge field and quasiperiodic onsite modulation. In the nearest-neighbor limit, we identify an anomalous transition from a fully erratic non-Hermitian skin effect (ENHSE) phase to a fully localized phase. Although the fractal dimension vanishes in both regimes, the Lyapunov exponent and the fluctuation of the eigenstate center of mass sharply distinguish them. For generic finite-size realizations, this transition is further accompanied by a complex-to-real spectral change under periodic boundary conditions and a change of spectral winding from nontrivial to trivial. With weak next-nearest-neighbor hopping, we uncover an anomalous mobility edge at the same location as in the Hermitian generalized Aubry-André-Harper model, but separating Anderson-localized states from ENHSE-type macroscopic-accumulation states rather than extended states. We further show that this anomalous localization structure is reflected in spectral winding and wave-packet dynamics: single realizations exhibit winding-dependent drift, winding-resolved averaging preserves opposite directional responses, and full disorder averaging largely restores Hermitian-like transport. Our results establish practical diagnostics of anomalous localization and mobility edges in non-Hermitian quasicrystals.

cond-mat.dis-nn↗