arXiv subjects

Hao Zhong

Publications and source records attributed to Hao Zhong.

At least 19 recordsLinked to original sources

arXiv2026

One-Step Epitaxial Access to Rhombohedral Graphene Flat-Band States on Step-Bunched SiC

Rhombohedral graphene multilayers provide a moiré-free platform for correlated and topological flat-band physics, but direct, transfer-free epitaxial access to thickness-tunable multilayers remains limited. Here we report a one-step graphitization route on 4$^\circ$ off-axis 4H-SiC, in which high-temperature flash annealing simultaneously drives self-organized step bunching and multilayer graphene formation. Atomic-resolution cross-sectional scanning transmission electron microscopy identify local ABC registry and distinguish rhombohedral from Bernal stacking. The thickness is tuned from bilayer to more than twenty layers by varying single parameter, the annealing temperature. Angle-resolved photoemission spectroscopy directly tracks the thickness-dependent evolution from interface-dominated low-energy states toward pronounced near-Fermi-level flat-band spectral weight in thick multilayers. Low-temperature scanning tunneling microscopy and spectroscopy on a 17-layer film further reveal a 13.4 meV low-energy spectral reconstruction and a $\sqrt{3} \times \sqrt{3}$ Kekulé-like modulation, providing microscopic signatures consistent with an intervalley-mixed electronic texture. This one-step, transfer-free approach establishes step-bunched SiC as an epitaxial platform that links stacking engineering with moiré-free correlated flat-band electronic states.

cond-mat.mtrl-sci↗

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

Higher-Order Congruence for Reciprocal Power Sums and Generalized Lehmer-Type Products

This paper investigates high-order congruences of reciprocal power sums and Lehmer-type products. Let $n\geq 1$ with $(n,6)=1$ and $e\in\{2,3,4,6\}$. For the reciprocal square sums \begin{equation*} S(n)=\sum_{\substack{r=1 \\ (r,n)=1}}^{\lfloor n/e \rfloor}\frac{1}{r^2} \end{equation*} we already know the form of the congruence modulo $n$. In this paper, motivated by the known congruences, we first extend these results to certain reciprocal sums of odd order and establish a uniform congruence modulo $n$ for \begin{equation*} S_m(n)=\sum_{\substack{r=1 \\ (r,n)=1}}^{\lfloor n/e \rfloor}\frac{1}{r^m} \end{equation*} We then study the generalized Lehmer-type product \begin{equation*} \prod_{d \mid n}\binom{kd-1}{\lfloor d/e \rfloor}^{μ(n/d)} \end{equation*} Although congruences modulo $n^3$ for this product have previously been obtained, higher-order congruences do not admit a comparably simple closed form. To address this difficulty, we derive an explicit truncated expansion in terms of complete exponential Bell polynomials. The results provide a unified framework for explicit computation and algorithmic verification of higher-order congruences involving reciprocal sums and related product expressions.

math.NT↗

arXiv2026

ACTIVE-o3: Empowering MLLMs with Active Perception via Pure Reinforcement Learning

Active vision, also known as active perception, refers to actively selecting where and how to look in order to gather task-relevant information. It is a critical component of efficient perception and decision-making in humans and advanced embodied agents. With the rise of Multimodal Large Language Models (MLLMs) as central planners in robotic systems, the lack of methods for equipping MLLMs with active perception has become a key gap. We first provide a systematic definition of MLLM-based active perception tasks and show that GPT-o3's zoom-in strategy can be viewed as a special case, though it suffers from low efficiency and inaccurate region selection. To address these issues, we propose ACTIVE-o3, a reinforcement learning framework built on GRPO that equips MLLMs with active perception capabilities. Leveraging a modular sensing-action design and a dual-form reward, ACTIVE-o3 autonomously learns efficient and stable region selection strategies without explicit region-selection supervision. We further establish a comprehensive benchmark covering both open-world tasks, including small- and dense-object grounding, and domain-specific scenarios, including remote sensing, autonomous driving, and interactive segmentation. Experimental results demonstrate that ACTIVE-o3 significantly enhances active perception capabilities compared to baselines. Moreover, we show that our framework not only preserves the model's general understanding ability but can also serve as a proxy task for leveraging perception data, further improving performance on benchmarks such as RealWorldQA and MME.

cs.CV↗

arXiv2026

Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching

Wide-baseline matching (WBM) requires integrating geometric understanding, viewpoint changes, fine-grained perception, and occlusion reasoning, making it a challenging testbed for spatial reasoning in multimodal large language models (MLLMs) deployed in physical environments. However, current MLLMs lack systematic evaluation and training frameworks for these capabilities. We introduce ReasonMatch-Bench, a benchmark stratified by viewpoint displacement and matching granularity across indoor, outdoor, and object-centric scenarios, and show that current MLLMs still struggle with fine-grained wide-baseline correspondence: on a difficult 90-sample subset, human annotators achieve 84.0 F1, while the best existing baseline reaches 37.2. To bridge this gap, we build a scalable data-generation pipeline that automatically extracts wide-baseline view pairs from large-scale video-3D corpora, including RGB-D videos and SfM reconstructions, yielding diverse and verifiable supervision. We further propose Dynamic Correspondence Reinforcement Learning (DCRL), which combines Image-Level Viewpoint Progression and Point-Level Correspondence Curriculum to improve WBM training through verifiable rewards without explicit CoT supervision. Extensive experiments show that DCRL substantially improves ReasonMatch-Bench and transfers to related spatial benchmarks, while maintaining general visual understanding performance with modest gains on several benchmarks.

cs.CV↗

arXiv2026

NoTVLA: Semantics-Preserving Robot Adaptation via Narrative Action Interfaces

Vision-Language-Action (VLA) models represent a pivotal advance in embodied intelligence, yet they confront critical barriers to real-world deployment, most notably catastrophic forgetting. This issue stems from their overreliance on continuous action sequences or action chunks, which inadvertently create isolated data silos that disrupt knowledge retention across tasks. To tackle these challenges, we propose the Narrowing of Trajectory VLA (NoTVLA) framework: a novel approach that narrows its focus to sparse trajectories, thereby avoiding the catastrophic forgetting associated with dense trajectory fine-tuning. A key innovation of NoTVLA lies in its trajectory planning strategy: instead of centering on the target object's trajectory, it leverages temporal compression and spatial reasoning pruning specifically for the robot end effector's trajectory. Furthermore, training is conducted using these sparse trajectories rather than dense action trajectories, an optimization that delivers remarkable practical advantages with better performance in zero-shot. In multi-task evaluation scenarios, NoTVLA achieves superior performance and generalization compared to pi0 while operating under two critical constraints: it uses over an order of magnitude less computing power than pi0 and requires no wrist-mounted camera. This design ensures that NoTVLA's operational accuracy closely approximates that of single-task expert models. Crucially, it also preserves the model's inherent language capabilities, enabling zero-shot generalization in specific scenarios, supporting unified model deployment across multiple robot platforms, and fostering a degree of generalization even when perceiving tasks from novel perspectives.

cs.RO↗

arXiv2026

Five-Minute Rule 40 Years Later: A First-Principles Revisit for Modern Memory Hierarchy

In 1987, Jim Gray and Gianfranco Putzolu introduced the five-minute rule, a simple, storage-memory-economics-based heuristic for deciding when data should live in DRAM rather than on storage. Subsequent revisits to the rule largely retained that economics-only view, leaving host costs, feasibility limits, and workload behavior out of scope. This paper revisits the rule from first principles, integrating host costs, DRAM bandwidth/capacity, and physics-grounded models of SSD performance and cost, and then embedding these elements in a constraint- and workload-aware framework that yields actionable provisioning guidance. We show that, for modern AI platforms, especially GPU-centric hosts paired with ultra-high-IOPS SSDs engineered for fine-grained random access, the DRAM$\leftrightarrow$flash caching threshold collapses from minutes to a few seconds. This shift reframes NAND flash memory as an \emph{active data tier} and exposes a broad research space across the hardware-software stack. We further introduce MQSim-Next, a calibrated SSD simulator that supports validation and sensitivity analysis and facilitates future architectural and system research. Finally, we present two concrete case studies that showcase the software system design space opened by such memory hierarchy paradigm shift. Overall, we turn a classical heuristic into an actionable, feasibility-aware analysis and provisioning framework and set the stage for further research on AI-era memory hierarchy.

cs.AR↗

arXiv2026

Exploring Spatial Intelligence from a Generative Perspective

Spatial intelligence is essential for multimodal large language models, yet current benchmarks largely assess it only from an understanding perspective. We ask whether modern generative or unified multimodal models also possess generative spatial intelligence (GSI), the ability to respect and manipulate 3D spatial constraints during image generation, and whether such capability can be measured or improved. We introduce GSI-Bench, the first benchmark designed to quantify GSI through spatially grounded image editing. It consists of two complementary components: GSI-Real, a high-quality real-world dataset built via a 3D-prior-guided generation and filtering pipeline, and GSI-Syn, a large-scale synthetic benchmark with controllable spatial operations and fully automated labeling. Together with a unified evaluation protocol, GSI-Bench enables scalable, model-agnostic assessment of spatial compliance and editing fidelity. Experiments show that fine-tuning unified multimodal models on GSI-Syn yields substantial gains on both synthetic and real tasks and, strikingly, also improves downstream spatial understanding. This provides the first clear evidence that generative training can tangibly strengthen spatial reasoning, establishing a new pathway for advancing spatial intelligence in multimodal models.

cs.CV↗

arXiv2026

OmniJigsaw: Enhancing Omni-Modal Reasoning via Modality-Orchestrated Reordering

To extend the reinforcement learning post-training paradigm to omni-modal models for concurrently bolstering video-audio understanding and collaborative reasoning, we propose OmniJigsaw, a generic self-supervised framework built upon a temporal reordering proxy task. Centered on the chronological reconstruction of shuffled audio-visual clips, this paradigm strategically orchestrates visual and auditory signals to compel cross-modal integration through three distinct strategies: Joint Modality Integration, Sample-level Modality Selection, and Clip-level Modality Masking. Recognizing that the efficacy of such proxy tasks is fundamentally tied to puzzle quality, we design a two-stage coarse-to-fine data filtering pipeline, which facilitates the efficient adaptation of OmniJigsaw to massive unannotated omni-modal data. Our analysis reveals a ``bi-modal shortcut phenomenon'' in joint modality integration and demonstrates that fine-grained clip-level modality masking mitigates this issue while outperforming sample-level modality selection. Extensive evaluations on 15 benchmarks show substantial gains in video, audio, and collaborative reasoning, validating OmniJigsaw as a scalable paradigm for self-supervised omni-modal learning.

cs.CV↗

arXiv2026

Preserving Source Video Realism: High-Fidelity Face Swapping for Cinematic Quality

Video face swapping is crucial in film and entertainment production, where achieving high fidelity and temporal consistency over long and complex video sequences remains a significant challenge. Inspired by recent advances in reference-guided image editing, we explore whether rich visual attributes from source videos can be similarly leveraged to enhance both fidelity and temporal coherence in video face swapping. Building on this insight, this work presents LivingSwap, the first video reference guided face swapping model. Our approach employs keyframes as conditioning signals to inject the target identity, enabling flexible and controllable editing. By combining keyframe conditioning with video reference guidance, the model performs temporal stitching to ensure stable identity preservation and high-fidelity reconstruction across long video sequences. To address the scarcity of data for reference-guided training, we construct a paired face-swapping dataset, Face2Face, and further reverse the data pairs to ensure reliable ground-truth supervision. Extensive experiments demonstrate that our method achieves state-of-the-art results, seamlessly integrating the target identity with the source video's expressions, lighting, and motion, while significantly reducing manual effort in production workflows. Project webpage: https://aim-uofa.github.io/LivingSwap

cs.CV↗

arXiv2026

PanoDP: Learning Collision-Free Navigation with Panoramic Depth and Differentiable Physics

Autonomous collision-free navigation in cluttered environments requires safe decision-making under partial observability with both static structure and dynamic obstacles. We present \textbf{PanoDP}, a communication-free learning framework that combines four-view panoramic depth perception with differentiable-physics-based training signals. PanoDP encodes panoramic depth using a lightweight CNN and optimizes policies with dense differentiable collision and motion-feasibility terms, improving training stability beyond sparse terminal collisions. We evaluate PanoDP on a controlled ring-to-center benchmark with systematic sweeps over agent count, obstacle density/layout, and dynamic behaviors, and further test out-of-distribution generalization in an external simulator (e.g., AirSim). Across settings, PanoDP increases collision-free and completion rates over single-view and non-physics-guided baselines under matched training budgets, and ablations (view masking, rotation augmentation) confirm the policy leverages 360-degree information. Code will be open source upon acceptance.

cs.RO↗

arXiv2026

LLaDA2.1: Speeding Up Text Diffusion via Token Editing

While LLaDA2.0 showcased the scaling potential of 100B-level block-diffusion models and their inherent parallelization, the delicate equilibrium between decoding speed and generation quality has remained an elusive frontier. Today, we unveil LLaDA2.1, a paradigm shift designed to transcend this trade-off. By seamlessly weaving Token-to-Token (T2T) editing into the conventional Mask-to-Token (M2T) scheme, we introduce a joint, configurable threshold-decoding scheme. This structural innovation gives rise to two distinct personas: the Speedy Mode (S Mode), which audaciously lowers the M2T threshold to bypass traditional constraints while relying on T2T to refine the output; and the Quality Mode (Q Mode), which leans into conservative thresholds to secure superior benchmark performances with manageable efficiency degrade. Furthering this evolution, underpinned by an expansive context window, we implement the first large-scale Reinforcement Learning (RL) framework specifically tailored for dLLMs, anchored by specialized techniques for stable gradient estimation. This alignment not only sharpens reasoning precision but also elevates instruction-following fidelity, bridging the chasm between diffusion dynamics and complex human intent. We culminate this work by releasing LLaDA2.1-Mini (16B) and LLaDA2.1-Flash (100B). Across 33 rigorous benchmarks, LLaDA2.1 delivers strong task performance and lightning-fast decoding speed. Despite its 100B volume, on coding tasks it attains an astounding 892 TPS on HumanEval+, 801 TPS on BigCodeBench, and 663 TPS on LiveCodeBench.

cs.LG↗

arXiv2026

SpatCode: Rotary-based Unified Encoding Framework for Efficient Spatiotemporal Vector Retrieval

Spatiotemporal vector retrieval has emerged as a critical paradigm in modern information retrieval, enabling efficient access to massive, heterogeneous data that evolve over both time and space. However, existing spatiotemporal retrieval methods are often extensions of conventional vector search systems that rely on external filters or specialized indices to incorporate temporal and spatial constraints, leading to inefficiency, architectural complexity, and limited flexibility in handling heterogeneous modalities. To overcome these challenges, we present a unified spatiotemporal vector retrieval framework that integrates temporal, spatial, and semantic cues within a coherent similarity space while maintaining scalability and adaptability to continuous data streams. Specifically, we propose (1) a Rotary-based Unified Encoding Method that embeds time and location into rotational position vectors for consistent spatiotemporal representation; (2) a Circular Incremental Update Mechanism that supports efficient sliding-window updates without global re-encoding or index reconstruction; and (3) a Weighted Interest-based Retrieval Algorithm that adaptively balances modality weights for context-aware and personalized retrieval. Extensive experiments across multiple real-world datasets demonstrate that our framework substantially outperforms state-of-the-art baselines in both retrieval accuracy and efficiency, while maintaining robustness under dynamic data evolution. These results highlight the effectiveness and practicality of the proposed approach for scalable spatiotemporal information retrieval in intelligent systems.

cs.IR↗

arXiv2026

CatchAll: Repository-Aware Exception Handling with Knowledge-Guided LLMs

Exception handling is a vital forward error-recovery mechanism in many programming languages, enabling developers to manage runtime anomalies through structured constructs (e.g., try-catch blocks). Improper or missing exception handling often leads to severe consequences, including system crashes and resource leaks. While large language models (LLMs) have demonstrated strong capabilities in code generation, they struggle with exception handling at the repository level, due to complex dependencies and contextual constraints. In this work, we propose CatchAll, a novel LLM-based approach for repository-aware exception handling. CatchAll equips LLMs with three complementary layers of exception-handling knowledge: (1) API-level exception knowledge, obtained from an empirically constructed API-exception mapping that characterizes the exception-throwing behaviors of APIs in real-world codebases; (2) repository-level execution context, which captures exception propagation by modeling contextual call traces around the target code; and (3) cross-repository handling knowledge, distilled from reusable exception-handling patterns mined from historical code across projects. The knowledge is encoded into structured prompts to guide the LLM in generating accurate and context-aware exception-handling code. To evaluate CatchAll, we construct two new benchmarks for repository-aware exception handling: a large-scale dataset RepoExEval and an executable subset RepoExEval-Exec. Experiments demonstrate that RepoExEval consistently outperforms state-of-the-art baselines, achieving a CodeBLEU score of 0.31 (vs. 0.27% for the best baseline), intent prediction accuracy of 60.1% (vs. 48.0%), and Pass@1 of 29% (vs. 25%). These results affirm RepoExEval's effectiveness in real-world repository-level exception handling.

cs.SE↗

arXiv2025

Quasi-steady electron-excitonic complexes coupling in a two-dimensional semiconductor

Excitons and their complexes govern optical-related behaviors in semiconductors. Here, using angle-resolved photoemission spectroscopy (ARPES), we have elucidated the light-matter interaction mediated by quasi-steady excitonic complexes within a monolayer of the prototypical two-dimensional (2D) semiconductor WSe2. Under continuous incident light, we have observed the generation of quasi-steady excitons and their complexes, encompassing ground and excited state excitons, trions, as well as their intricate interplay. We further show spectral evidence of electronic excitation states within the background of quasi-steady excitonic complexes, characterized by valence band (VB) effective mass renormalization, the enhanced spin-orbit coupling (SOC), the formation of an excitonic gap near the Fermi level (EF ) of the conduction band (CB), and intervalley excitonic band folding. Our findings not only unveil a quasi-steady excitonic complex background for the creation of diverse electronic excitations in 2D semiconductors but also offer new insights into the role of excitons in the charge density wave (CDW) formation mechanism and facilitate the advancement of correlated electronic state engineering based on the coupling between electrons and excitonic complexes in a quasi-equilibrium state.

cond-mat.str-el↗

arXiv2025

Surface band-selective moiré effect induces flat band in mixed-dimensional heterostructures

In this work, we reveal a curious type of moiré effect that selectively modifies the surface states of bulk crystal. We synthesize mixed-dimensional heterostructures consisting of a noble gas monolayer grow on the surface of bulk Bi(111), and determine the electronic structure of the heterostructures using angle-resolved photoemission spectroscopy. We directly observe moiré replicas of the Bi(111) surface states, while the bulk states remain barely changed. Meanwhile, we achieve control over the moiré period in the range of 25 Å to 80 Å by selecting monolayers of different noble gases and adjusting the annealing temperature. At large moiré periods, we observe hybridization between the surface band replicas, which leads to the formation of a correlated flat band. Our results serve as a bridge for understanding the moiré modulation effect from 2D to 3D systems, and provide a feasible approach for the realization of correlated phenomena through the engineering of surface states via moiré effects.

cond-mat.str-el↗

arXiv2025

GUI-Shepherd: Reliable Process Reward and Verification for Long-Sequence GUI Tasks

Autonomous agents for long-sequence Graphical User Interface tasks are hindered by sparse rewards and the intractable credit assignment problem. To address these challenges, we introduce GUI-Shepherd, a Process Reward Model that provides dense, step-by-step feedback to guide agents. GUI-Shepherd is trained on a diverse large-scale data set of $52$k interactions that features human-annotated scores and GPT-4o generated rationales, enabling it to serve both as a reward provider for RL training and as a verifier for inference. As far as we know, we are the first to conduct a systematic study of process supervision in GUI agents, across diverse settings from online long-horizon tasks to offline single-step prediction. On the online AndroidWorld benchmark, GUI-Shepherd improves success rate by $7.7$ points via multi-turn online PPO, significantly outperforming Outcome Reward Model based competitors. When used as an inference verifier, it brings $5.1$ points improvements. The benefits generalize to the offline AndroidControl benchmark, with gains of $2.2$ points as a reward provider and $4.3$ points as a verifier. Collectively, our results establish that high-fidelity process supervision is critical for building more capable GUI agents and present a generalizable solution.

cs.AI↗

arXiv2025

$q$-Binomial Identities Finder

This paper presents a symbolic computation method for automatically transforming $q$-hypergeometric identities to $q$-binomial identities. Through this method, many previously proven $q$-binomial identities, including $q$-Saalschütz's formula and $q$-Suranyi's formula, are re-fund, and numerous new ones are discovered. Moreover, the generation of the identities is accompanied by the corresponding proofs. During the transformation process, different ranges of variable values and various combinations of $q$-Pochhammer symbols yield different identities. The algorithm maps variable constraints to positive elements in an ordered vector space and employs a backtracking method to provide the feasible variable constraints and $q$-binomial coefficient combinations for each step.

math.CO↗