arXiv subjects

Bo Zhao

Publications and source records attributed to Bo Zhao.

At least 109 records · Page 6Linked to original sources

arXiv2026

Mode Conversion of Hyperbolic Phonon Polaritons in van der Waals terraces

Electromagnetic hyperbolicity has driven key functionalities in nanophotonics, including super-resolution imaging, efficient energy control, and extreme light manipulation. Central to these advances are hyperbolic polaritons - nanometer-scale light-matter waves - spanning multiple energy-momentum dispersion orders with distinct mode profiles and incrementally high optical momenta. In this work, we report the mode conversion of hyperbolic polaritons across different dispersion orders by breaking the structure symmetry in engineered step-shape van der Waals (vdW) terraces. The mode conversion from the fundamental to high-order hyperbolic polaritons is imaged using scattering-type scanning near-field optical microscopy (s-SNOM) on both hexagonal boron nitride (hBN) and alpha-phase molybdenum trioxide (alpha-MoO3) vdW terraces. Our s-SNOM data, augmented with electromagnetic simulations, further demonstrate the alteration of polariton mode conversion by varying the step size of vdW terraces. The mode conversion reported here offers a practical approach toward integrating previously independent different-order hyperbolic polaritons with ultra-high momenta, paving the way for promising applications in nano-optical circuits, sensing, computation, information processing, and super-resolution imaging.

physics.optics↗

arXiv2026

Efficient Multimodal Large Language Models: A Survey

In the past year, Multimodal Large Language Models (MLLMs) have demonstrated remarkable performance in tasks such as visual question answering, visual understanding and reasoning. However, the extensive model size and high training and inference costs have hindered the widespread application of MLLMs in academia and industry. Thus, studying efficient and lightweight MLLMs has enormous potential, especially in edge computing scenarios. In this survey, we provide a comprehensive and systematic review of the current state of efficient MLLMs. Specifically, we summarize the timeline of representative efficient MLLMs, research state of efficient structures and strategies, and the applications. Finally, we discuss the limitations of current efficient MLLM research and promising future directions. Please refer to our GitHub repository for more details: https://github.com/lijiannuist/Efficient-Multimodal-LLMs-Survey.

cs.CV↗

arXiv2026

MomentSeeker: A Task-Oriented Benchmark For Long-Video Moment Retrieval

Accurately locating key moments within long videos is crucial for solving long video understanding (LVU) tasks. However, existing benchmarks are either severely limited in terms of video length and task diversity, or they focus solely on the end-to-end LVU performance, making them inappropriate for evaluating whether key moments can be accurately accessed. To address this challenge, we propose MomentSeeker, a novel benchmark for long-video moment retrieval (LMVR), distinguished by the following features. First, it is created based on long and diverse videos, averaging over 1200 seconds in duration and collected from various domains, e.g., movie, anomaly, egocentric, and sports. Second, it covers a variety of real-world scenarios in three levels: global-level, event-level, object-level, covering common tasks like action recognition, object localization, and causal reasoning, etc. Third, it incorporates rich forms of queries, including text-only queries, image-conditioned queries, and video-conditioned queries. On top of MomentSeeker, we conduct comprehensive experiments for both generation-based approaches (directly using MLLMs) and retrieval-based approaches (leveraging video retrievers). Our results reveal the significant challenges in long-video moment retrieval in terms of accuracy and efficiency, despite improvements from the latest long-video MLLMs and task-specific fine-tuning. We have publicly released MomentSeeker(https://yhy-2000.github.io/MomentSeeker/) to facilitate future research in this area.

cs.CV↗

arXiv2025

Achieving high-performance polarization-independent nonreciprocal thermal radiation with pattern-free heterostructures

Many advanced energy harvesting technologies rely on advanced control of thermal emission. Recently, it has been shown that the emissivity and absorptivity of thermal emitters can be controlled independently in nonreciprocal emitters. While significant progress has been made in engineering these nonreciprocal thermal emitters, realizing a highly efficient, pattern-free emitter capable of supporting dual-polarization nonreciprocal emission remains a challenging task. Existing solutions are largely based on metamaterials and exhibit polarization-dependent behavior. This work proposes pattern-free multilayer heterostructures combining magneto-optical and magnetic Weyl semimetal materials and systematically evaluates their nonreciprocal emission performance for p- and s-polarized waves. The findings show that omnidirectional polarization-independent nonreciprocity can be achieved utilizing multilayer structures with different magnetization directions that do not follow simple vector summation. To further enhance the performance, Pareto optimization is employed to tune the key design parameters, enabling the maximization of nonreciprocal thermal emission in a given wavelength range. This approach offers a versatile strategy for designing high-performance thermal emitters tailored for multi-objective optical functionalities.

physics.optics↗

arXiv2025

Symmetry in Neural Network Parameter Spaces

Modern deep learning models are highly overparameterized, resulting in large sets of parameter configurations that yield the same outputs. A significant portion of this redundancy is explained by symmetries in the parameter space--transformations that leave the network function unchanged. These symmetries shape the loss landscape and constrain learning dynamics, offering a new lens for understanding optimization, generalization, and model complexity that complements existing theory of deep learning. This survey provides an overview of parameter space symmetry. We summarize existing literature, uncover connections between symmetry and learning theory, and identify gaps and opportunities in this emerging field.

cs.LG↗

arXiv2025

TimeScope: Towards Task-Oriented Temporal Grounding In Long Videos

Identifying key temporal intervals within long videos, known as temporal grounding (TG), is important to video understanding and reasoning tasks. In this paper, we introduce a new form of the temporal grounding problem, \textbf{Task-oriented Temporal Grounding} (\textbf{ToTG}), which is driven by the requirements of downstream tasks rather than explicit time-interval descriptions. For example, a ToTG input may be "explain why the man in the video is sent to the hospital," whereas traditional TG would take an explicit temporal description such as "the moments when the man is tripped by a stone and falls to the ground." This new ToTG formulation presents significant challenges for existing TG methods, as it requires jointly performing deep task comprehension and fine-grained temporal localization within long videos. To address these challenges, we conduct a systematic set of studies. First, we construct \textbf{a new benchmark ToTG-Bench}, which comprehensively evaluates ToTG performance across diverse settings. Second, we introduce \textbf{a new temporal-ground method TimeScope}, which performs coarse-to-fine localization through a progressive reasoning process. Leveraging extensive supervised fine-tuning with carefully curated chain-of-thought (CoT) data from a variety of scenarios, TimeScope generalizes effectively across tasks and domains. Our evaluation demonstrates \textbf{TimeScope's empirical advantages} over existing baselines from three perspectives: (1) substantial improvements in grounding precision, (2) significant benefits to downstream tasks, and (3) strong generalizability across different scenarios. All models, datasets, and source code will be fully open-sourced to support future research in this area.

cs.CV↗

arXiv2025

Evo-1: Lightweight Vision-Language-Action Model with Preserved Semantic Alignment

Vision-Language-Action (VLA) models have emerged as a powerful framework that unifies perception, language, and control, enabling robots to perform diverse tasks through multimodal understanding. However, current VLA models typically contain massive parameters and rely heavily on large-scale robot data pretraining, leading to high computational costs during training, as well as limited deployability for real-time inference. Moreover, most training paradigms often degrade the perceptual representations of the vision-language backbone, resulting in overfitting and poor generalization to downstream tasks. In this work, we present Evo-1, a lightweight VLA model that reduces computation and improves deployment efficiency, while maintaining strong performance without pretraining on robot data. Evo-1 builds on a native multimodal Vision-Language model (VLM), incorporating a novel cross-modulated diffusion transformer along with an optimized integration module, together forming an effective architecture. We further introduce a two-stage training paradigm that progressively aligns action with perception, preserving the representations of the VLM. Notably, with only 0.77 billion parameters, Evo-1 achieves state-of-the-art results on the Meta-World and RoboTwin suite, surpassing the previous best models by 12.4% and 6.9%, respectively, and also attains a competitive result of 94.8% on LIBERO. In real-world evaluations, Evo-1 attains a 78% success rate with high inference frequency and low memory overhead, outperforming all baseline methods. We release code, data, and model weights to facilitate future research on lightweight and efficient VLA models.

cs.RO↗

arXiv2025

Evo-0: Vision-Language-Action Model with Implicit Spatial Understanding

Vision-Language-Action (VLA) models have emerged as a promising framework for enabling generalist robots capable of perceiving, reasoning, and acting in the real world. These models usually build upon pretrained Vision-Language Models (VLMs), which excel at semantic understanding due to large-scale image and text pretraining. However, existing VLMs typically lack precise spatial understanding capabilities, as they are primarily tuned on 2D image-text pairs without 3D supervision. To address this limitation, recent approaches have incorporated explicit 3D inputs such as point clouds or depth maps, but this necessitates additional depth sensors or pre-trained depth estimation models, which may yield defective results. In contrast, our work introduces a plug-and-play module that implicitly incorporates 3D geometry features into VLA models by leveraging an off-the-shelf visual geometry foundation model. This integration provides the model with depth-aware visual representations, improving its ability to understand the geometric structure of the scene and the spatial relationships among objects from RGB images alone. We evaluate our method on a set of spatially challenging tasks in both simulation and the real world. Extensive evaluations show that our method significantly improves the performance of state-of-the-art VLA models across diverse scenarios.

cs.RO↗

arXiv2025

Mask2IV: Interaction-Centric Video Generation via Mask Trajectories

Generating interaction-centric videos, such as those depicting humans or robots interacting with objects, is crucial for embodied intelligence, as they provide rich and diverse visual priors for robot learning, manipulation policy training, and affordance reasoning. However, existing methods often struggle to model such complex and dynamic interactions. While recent studies show that masks can serve as effective control signals and enhance generation quality, obtaining dense and precise mask annotations remains a major challenge for real-world use. To overcome this limitation, we introduce Mask2IV, a novel framework specifically designed for interaction-centric video generation. It adopts a decoupled two-stage pipeline that first predicts plausible motion trajectories for both actor and object, then generates a video conditioned on these trajectories. This design eliminates the need for dense mask inputs from users while preserving the flexibility to manipulate the interaction process. Furthermore, Mask2IV supports versatile and intuitive control, allowing users to specify the target object of interaction and guide the motion trajectory through action descriptions or spatial position cues. To support systematic training and evaluation, we curate two benchmarks covering diverse action and object categories across both human-object interaction and robotic manipulation scenarios. Extensive experiments demonstrate that our method achieves superior visual realism and controllability compared to existing baselines.

cs.CV↗

arXiv2025

Ferroelectricity-driven altermagnetism in two-dimensional van der Waals multiferroics

Altermagnets (AMs) are a recently identified class of unconventional collinear compensated antiferromagnets that exhibit momentum-dependent spin splitting despite having zero net magnetization. This unconventional magnetic order gives rise to a range of phenomena, including the anomalous Hall effect, chiral magnons, and nonlinear photocurrents. Here, using spin space group (SSG) symmetry analysis and first-principles calculations, we demonstrate an efficient strategy to control altermagnetism in two-dimensional multiferroics through ferroelectric polarization and interlayer sliding. For material realization, we find that monolayer and bilayer FeCuP2S6 exhibit finite spin splitting when ferroelectric sublattices are connected by nonsymmorphic screw-axis operations rather than pure translation or inversion symmetry. Interlayer sliding further enables reversible switching or suppression of spin splitting through modifications of the SSG. Our calculations further reveal that the anomalous Hall response serves as a direct probe of these spin-split states. These findings establish two-dimensional van der Waals multiferroics as promising platforms for realizing electrically controllable altermagnetism and advancing next-generation spintronic and magnetoelectric technologies.

cond-mat.mtrl-sci↗

arXiv2025

ZX-DB: A Graph Database for Quantum Circuit Simplification and Rewriting via the ZX-Calculus

Quantum computing is an emerging computational paradigm with the potential to outperform classical computers in solving a variety of problems. To achieve this, quantum programs are typically represented as quantum circuits, which must be optimized and adapted for target hardware through quantum circuit compilation. We introduce ZX-DB, a data-driven system that performs quantum circuit simplification and rewriting inside a graph database using ZX-calculus, a complete graphical formalism for quantum mechanics. ZX-DB encodes ZX-calculus rewrite rules as standard openCypher queries and executes them on an example graph database engine, Memgraph, enabling efficient, database-native transformations of large-scale quantum circuits. ZX-DB integrates correctness validation via tensor and graph equivalence checks and is evaluated against the state-of-the-art PyZX framework. Experimental results show that ZX-DB achieves up to an order-of-magnitude speedup for independent rewrites, while exposing pattern-matching bottlenecks in current graph database engines. By uniting quantum compilation and graph data management, ZX-DB opens a new systems direction toward scalable, database-supported quantum computing pipelines.

quant-ph↗

arXiv2025

Compensating Distribution Drifts in Class-incremental Learning of Pre-trained Vision Transformers

Recent advances have shown that sequential fine-tuning (SeqFT) of pre-trained vision transformers (ViTs), followed by classifier refinement using approximate distributions of class features, can be an effective strategy for class-incremental learning (CIL). However, this approach is susceptible to distribution drift, caused by the sequential optimization of shared backbone parameters. This results in a mismatch between the distributions of the previously learned classes and that of the updater model, ultimately degrading the effectiveness of classifier performance over time. To address this issue, we introduce a latent space transition operator and propose Sequential Learning with Drift Compensation (SLDC). SLDC aims to align feature distributions across tasks to mitigate the impact of drift. First, we present a linear variant of SLDC, which learns a linear operator by solving a regularized least-squares problem that maps features before and after fine-tuning. Next, we extend this with a weakly nonlinear SLDC variant, which assumes that the ideal transition operator lies between purely linear and fully nonlinear transformations. This is implemented using learnable, weakly nonlinear mappings that balance flexibility and generalization. To further reduce representation drift, we apply knowledge distillation (KD) in both algorithmic variants. Extensive experiments on standard CIL benchmarks demonstrate that SLDC significantly improves the performance of SeqFT. Notably, by combining KD to address representation drift with SLDC to compensate distribution drift, SeqFT achieves performance comparable to joint training across all evaluated datasets. Code: https://github.com/raoxuan98-hash/sldc.git.

cs.CV↗

arXiv2025

Remanent crustal strain on Mars in non-poikilitic olivine of NWA 7721

We present a multiscale microstructural analysis of olivine from the non-poikilitic lithology of the poikilitic shergottite NWA 7721, using dark-field X-ray microscopy (DFXM), electron backscatter diffraction (EBSD), and context in situ 2D micro-XRD. A single olivine crystal contains two distinct subgrain populations. Type 1 subgrains are fine (1-5 micrometers), randomly oriented, and nearly strain-free, whereas Type 2 subgrains are coarse (greater than 30 micrometers), aligned, and strongly strained. Layered DFXM data reveal slip-band features in Type 2 that are absent in Type 1. We interpret Type 1 as products of shock-induced recrystallization, whereas Type 2 preserves remnants of a highly deformed parent grain. This bimodal microstructure, not observed in other Martian meteorites including the paired NWA 1950 and ALH A77005, points to a heterogeneous response to impact influenced by pre-existing strain in the olivine grain. We propose that NWA 7721 olivine experienced substantial crustal or magmatic stress before impact. The subsequent shock wave imposed a rapid load-release cycle that mobilized dislocations and produced low-angle boundaries in Type 2, while driving recrystallization of Type 1. Grain-growth constraints limit the post-shock heating duration to approximately 2.3 s, consistent with rapid quenching. These results provide the first evidence that non-poikilitic olivine in NWA 7721 preserves dynamic crustal deformation on Mars in the Late Amazonian.

astro-ph.EP↗

arXiv2025

Broad Feshbach resonance with a large background scattering length in a fermionic atom-molecule mixture

We report the observation of a broad magnetic Feshbach resonance with a large background scattering length in an ultracold fermionic mixture of $^{23}$Na$^{40}$K molecules and $^{40}$K atoms, with both species prepared in their lowest hyperfine states. The Feshbach resonance is characterized by measuring resonantly enhanced loss rates and elastic scattering cross sections via cross-species thermalization. The large background scattering length can drive the atom-molecule mixture into the hydrodynamic regime when the magnetic field is far from the resonance. We observe that the center-of-mass motions of the atoms and molecules are phase-locked and oscillate with a common frequency due to hydrodynamic drag effects. This broad atom-molecule Feshbach resonance with its large background scattering length opens up a new avenue towards studying strongly interacting fermionic gases with mass imbalance.

cond-mat.quant-gas↗

arXiv2025

SHARP: Shared State Reduction for Efficient Matching of Sequential Patterns

The detection of sequential patterns in data is a basic functionality of modern data processing systems for complex event processing (CEP), OLAP, and retrieval-augmented generation (RAG). In practice, pattern matching is challenging, since common applications rely on a large set of patterns that shall be evaluated with tight latency bounds. At the same time, matching needs to maintain state, i.e., intermediate results, that grows exponentially in the input size. Hence, systems turn to best-effort processing, striving for maximal recall under a latency bound. Existing techniques, however, consider each pattern in isolation, neglecting the optimization potential induced by state sharing in pattern matching. In this paper, we present SHARP, a library that employs state reduction to achieve efficient best-effort pattern matching. To this end, SHARP incorporates state sharing between patterns through a new abstraction, coined pattern-sharing degree (PSD). At runtime, this abstraction facilitates the categorization and indexing of partial pattern matches. Based thereon, once a latency bound is exceeded, SHARP realizes best-effort processing by selecting a subset of partial matches for further processing in constant time. In experiments with real-world data, SHARP achieves a recall of 97%, 96% and 73% for pattern matching in CEP, OLAP, and RAG applications, under a bound of 50% of the average processing latency.

cs.DB↗

arXiv2025

Optimizing Reasoning Efficiency through Prompt Difficulty Prediction

Reasoning language models perform well on complex tasks but are costly to deploy due to their size and long reasoning traces. We propose a routing approach that assigns each problem to the smallest model likely to solve it, reducing compute without sacrificing accuracy. Using intermediate representations from s1.1-32B, we train lightweight predictors of problem difficulty or model correctness to guide routing across a pool of reasoning models. On diverse math benchmarks, routing improves efficiency over random assignment and matches s1.1-32B's performance while using significantly less compute. Our results demonstrate that difficulty-aware routing is effective for cost-efficient deployment of reasoning models.

cs.LG↗

arXiv2025

UtilGen: Utility-Centric Generative Data Augmentation with Dual-Level Task Adaptation

Data augmentation using generative models has emerged as a powerful paradigm for enhancing performance in computer vision tasks. However, most existing augmentation approaches primarily focus on optimizing intrinsic data attributes -- such as fidelity and diversity -- to generate visually high-quality synthetic data, while often neglecting task-specific requirements. Yet, it is essential for data generators to account for the needs of downstream tasks, as training data requirements can vary significantly across different tasks and network architectures. To address these limitations, we propose UtilGen, a novel utility-centric data augmentation framework that adaptively optimizes the data generation process to produce task-specific, high-utility training data via downstream task feedback. Specifically, we first introduce a weight allocation network to evaluate the task-specific utility of each synthetic sample. Guided by these evaluations, UtilGen iteratively refines the data generation process using a dual-level optimization strategy to maximize the synthetic data utility: (1) model-level optimization tailors the generative model to the downstream task, and (2) instance-level optimization adjusts generation policies -- such as prompt embeddings and initial noise -- at each generation round. Extensive experiments on eight benchmark datasets of varying complexity and granularity demonstrate that UtilGen consistently achieves superior performance, with an average accuracy improvement of 3.87% over previous SOTA. Further analysis of data influence and distribution reveals that UtilGen produces more impactful and task-relevant synthetic data, validating the effectiveness of the paradigm shift from visual characteristics-centric to task utility-centric data augmentation.

cs.CV↗

arXiv2025

Boron Nitride Coatings for the Enhanced Detection of Neutrons in CR-39

The neutron detection efficiency of Columbia Resin 39 (CR-39) nuclear track detectors was assessed for AmBe, 252Cf, and D-T (14 MeV) neutron source spectra. A boron nitride (BN) coating for CR-39 was developed to enhance the neutron detection efficiency by converting neutrons into energetic alpha particles through the well-known 10B(n,a)7Li reaction. Separate partially coated CR-39 pieces were exposed to each neutron source and subsequently analyzed under optical microscope and through large-area Scanning Electron Microscopy (SEM) imaging over the irradiated area. The detection efficiency (tracks per neutron) was evaluated for each source spectra under optical and scanning electron microscopes and with or without BN coating. This resulted in a comprehensive guide to neutron detection with various sources using CR-39.

physics.ins-det↗