Search

This paper aims to reconstruct hundreds of people's 3D poses, shapes, and locations from a single image with unknown camera parameters. Due to the small and highly varying 2D human scales, depth ambiguity, and perspective distortion, no existing methods can achieve globally consistent reconstruction and accurate reprojection. To address these challenges, we first propose Crowd3D, which leverages a new concept, Human-scene Virtual Interaction Point (HVIP), to convert the complex 3D human localization into 2D-pixel localization with robust camera and ground estimation to achieve globally consistent reconstruction. To achieve stable generalization on different camera FoVs without test-time optimization, we propose an extended version, Crowd3D++, which eliminates the influence of camera parameters and the cropping operation by the proposed canonical upright space and ground-aware normalization transform. In the defined upright space, Crowd3D++ also designs an HVIPNet to regress 2D HVIP and infer the depths. Besides, we contribute two benchmark datasets, LargeCrowd and SyntheticCrowd, for evaluating crowd reconstruction in large scenes. The source code and data will be made publicly available after acceptance., Comment: 14 pages including reference

Geometric estimation is required for scene understanding and analysis in panoramic 360{\deg} images. Current methods usually predict a single feature, such as depth or surface normal. These methods can lack robustness, especially when dealing with intricate textures or complex object surfaces. We introduce a novel multi-task learning (MTL) network that simultaneously estimates depth and surface normals from 360{\deg} images. Our first innovation is our MTL architecture, which enhances predictions for both tasks by integrating geometric information from depth and surface normal estimation, enabling a deeper understanding of 3D scene structure. Another innovation is our fusion module, which bridges the two tasks, allowing the network to learn shared representations that improve accuracy and robustness. Experimental results demonstrate that our MTL architecture significantly outperforms state-of-the-art methods in both depth and surface normal estimation, showing superior performance in complex and diverse scenes. Our model's effectiveness and generalizability, particularly in handling intricate surface textures, establish it as a new benchmark in 360{\deg} image geometric estimation. The code and model are available at \url{https://github.com/huangkun101230/360MTLGeometricEstimation}., Comment: 18 pages, this paper is accepted by Computational Visual Media Journal (CVMJ) but not pushlished yet

The low-energy $J/\psi N$ scattering is important for various reasons: it is related to the hidden-charm $P_c$ pentaquark states, provides insights into the role of gluons in nucleon structures, and is relevant to the $J/\psi$ properties in nuclear medium. The scattering can happen through two distinct mechanisms: the coupled-channel mechanism via open-charm meson-baryon intermediate states, and the soft-gluon exchange mechanism. We investigate the $J/\psi N$ $S$-wave scattering length through both mechanisms, and find that the soft-gluon exchange mechanism leads to a scattering length at least one order of magnitude larger than that from the coupled-channel mechanism and thus is the predominant one. The findings can be verified by lattice calculations and will enhance our understanding of the scattering processes breaking the Okubo-Zweig-Iizuka rule.

Recent breakthroughs in artificial intelligence have driven a paradigm shift, where large language models (LLMs) with billions or trillions of parameters are trained on vast datasets, achieving unprecedented success across a series of language tasks. However, despite these successes, LLMs still rely on probabilistic modeling, which often captures spurious correlations rooted in linguistic patterns and social stereotypes, rather than the true causal relationships between entities and events. This limitation renders LLMs vulnerable to issues such as demographic biases, social stereotypes, and LLM hallucinations. These challenges highlight the urgent need to integrate causality into LLMs, moving beyond correlation-driven paradigms to build more reliable and ethically aligned AI systems. While many existing surveys and studies focus on utilizing prompt engineering to activate LLMs for causal knowledge or developing benchmarks to assess their causal reasoning abilities, most of these efforts rely on human intervention to activate pre-trained models. How to embed causality into the training process of LLMs and build more general and intelligent models remains unexplored. Recent research highlights that LLMs function as causal parrots, capable of reciting causal knowledge without truly understanding or applying it. These prompt-based methods are still limited to human interventional improvements. This survey aims to address this gap by exploring how causality can enhance LLMs at every stage of their lifecycle-from token embedding learning and foundation model training to fine-tuning, alignment, inference, and evaluation-paving the way for more interpretable, reliable, and causally-informed models. Additionally, we further outline six promising future directions to advance LLM development, enhance their causal reasoning capabilities, and address the current limitations these models face.

Discovering causal structures with latent variables from observational data is a fundamental challenge in causal discovery. Existing methods often rely on constraint-based, iterative discrete searches, limiting their scalability to large numbers of variables. Moreover, these methods frequently assume linearity or invertibility, restricting their applicability to real-world scenarios. We present new theoretical results on the identifiability of nonlinear latent hierarchical causal models, relaxing previous assumptions in literature about the deterministic nature of latent variables and exogenous noise. Building on these insights, we develop a novel differentiable causal discovery algorithm that efficiently estimates the structure of such models. To the best of our knowledge, this is the first work to propose a differentiable causal discovery method for nonlinear latent hierarchical models. Our approach outperforms existing methods in both accuracy and scalability. We demonstrate its practical utility by learning interpretable hierarchical latent structures from high-dimensional image data and demonstrate its effectiveness on downstream tasks., Comment: 25 pages with references, 7 figures

Closed-loop simulation is crucial for end-to-end autonomous driving. Existing sensor simulation methods (e.g., NeRF and 3DGS) reconstruct driving scenes based on conditions that closely mirror training data distributions. However, these methods struggle with rendering novel trajectories, such as lane changes. Recent works have demonstrated that integrating world model knowledge alleviates these issues. Despite their efficiency, these approaches still encounter difficulties in the accurate representation of more complex maneuvers, with multi-lane shifts being a notable example. Therefore, we introduce ReconDreamer, which enhances driving scene reconstruction through incremental integration of world model knowledge. Specifically, DriveRestorer is proposed to mitigate artifacts via online restoration. This is complemented by a progressive data update strategy designed to ensure high-quality rendering for more complex maneuvers. To the best of our knowledge, ReconDreamer is the first method to effectively render in large maneuvers. Experimental results demonstrate that ReconDreamer outperforms Street Gaussians in the NTA-IoU, NTL-IoU, and FID, with relative improvements by 24.87%, 6.72%, and 29.97%. Furthermore, ReconDreamer surpasses DriveDreamer4D with PVG during large maneuver rendering, as verified by a relative improvement of 195.87% in the NTA-IoU metric and a comprehensive user study., Comment: Project Page: https://recondreamer.github.io

Hydrogen-rich superconductors are promising candidates to achieve room-temperature superconductivity. However, the extreme pressures needed to stabilize these structures significantly limit their practical applications. An effective strategy to reduce the external pressure is to add a light element M that binds with H to form MHx units, acting as a chemical precompressor. We exemplify this idea by performing ab initio calculations of the Ac-Be-H phase diagram, proving that the metallization pressure of Ac-H binaries, for which critical temperatures as high as 200 K were predicted at 200 GPa, can be significantly reduced via beryllium incorporation. We identify three thermodynamically stable (AcBe2H10, AcBeH8, and AcBe2H14) and four metastable compounds (fcc AcBeH8, AcBeH10, AcBeH12 and AcBe2H16). All of them are superconductors. In particular, fcc AcBeH8 remains dynamically stable down to 10 GPa, where it exhibits a superconducting transition temperature Tc of 181 K. The Be-H bonds are responsible for the exceptional properties of these ternary compounds and allow them to remain dynamically stable close to ambient pressure. Our results suggest that high-Tc superconductivity in hydrides is achievable at low pressure and may stimulate experimental synthesis of ternary hydrides.

Altermagnets (AM) are a recently discovered third class of collinear magnets, and have been attracting significant interest in the field of condensed matter physics. Here, based on first-principles calculations and theoretical analysis, we propose four two-dimensional (2D) magnetic materials--monolayer V$_2$Te$_2$O, V$_2$STeO, V$_2$SSeO, and V$_2$S$_2$O--as candidates for altermagnetic materials. We show that these materials are semiconductors with spin-splitting in their nonrelativistic band structures. Furthermore, in the band structure, there are a pair of Dirac-type valleys located at the time-reversal invariant momenta (TRIM) X and Y points. These two valleys are connected by crystal symmetry instead of time-reversal symmetry. We investigate the strain effect on the band structure and find that uniaxial strain can induce valley polarization, topological states in these monolayer materials. Moreover, piezomagnetism can be realized upon finite doping. Our result reveals interesting valley physics in monolayer V$_2$Te$_2$O, V$_2$STeO, V$_2$SSeO, and V$_2$S$_2$O, suggesting their great potential for valleytronics, spintronics, and multifunctional nanoelectronics applications., Comment: 7 pages, 8 figures

There have been recent efforts to extend the Chain-of-Thought (CoT) paradigm to Multimodal Large Language Models (MLLMs) by finding visual clues in the input scene, advancing the visual reasoning ability of MLLMs. However, current approaches are specially designed for the tasks where clue finding plays a major role in the whole reasoning process, leading to the difficulty in handling complex visual scenes where clue finding does not actually simplify the whole reasoning task. To deal with this challenge, we propose a new visual reasoning paradigm enabling MLLMs to autonomously modify the input scene to new ones based on its reasoning status, such that CoT is reformulated as conducting simple closed-loop decision-making and reasoning steps under a sequence of imagined visual scenes, leading to natural and general CoT construction. To implement this paradigm, we introduce a novel plug-and-play imagination space, where MLLMs conduct visual modifications through operations like focus, ignore, and transform based on their native reasoning ability without specific training. We validate our approach through a benchmark spanning dense counting, simple jigsaw puzzle solving, and object placement, challenging the reasoning ability beyond clue finding. The results verify that while existing techniques fall short, our approach enables MLLMs to effectively reason step by step through autonomous imagination. Project page: https://future-item.github.io/autoimagine-site.

Large Language Models (LLMs) have demonstrated impressive capabilities across various tasks. However, LLMs often struggle with spatial reasoning which is one essential part of reasoning and inference and requires understanding complex relationships between objects in space. This paper proposes a novel neural-symbolic framework that enhances LLMs' spatial reasoning abilities. We evaluate our approach on two benchmark datasets: StepGame and SparQA, implementing three distinct strategies: (1) ASP (Answer Set Programming)-based symbolic reasoning, (2) LLM + ASP pipeline using DSPy, and (3) Fact + Logical rules. Our experiments demonstrate significant improvements over the baseline prompting methods, with accuracy increases of 40-50% on StepGame} dataset and 3-13% on the more complex SparQA dataset. The "LLM + ASP" pipeline achieves particularly strong results on the tasks of Finding Relations (FR) and Finding Block (FB) questions, though performance varies across different question types. The impressive results suggest that while neural-symbolic approaches offer promising directions for enhancing spatial reasoning in LLMs, their effectiveness depends heavily on the specific task characteristics and implementation strategies. We propose an integrated, simple yet effective set of strategies using a neural-symbolic pipeline to boost spatial reasoning abilities in LLMs. This pipeline and its strategies demonstrate strong and broader applicability to other reasoning domains in LLMs, such as temporal reasoning, deductive inference etc.

Multispectral object detection, utilizing RGB and TIR (thermal infrared) modalities, is widely recognized as a challenging task. It requires not only the effective extraction of features from both modalities and robust fusion strategies, but also the ability to address issues such as spectral discrepancies, spatial misalignment, and environmental dependencies between RGB and TIR images. These challenges significantly hinder the generalization of multispectral detection systems across diverse scenarios. Although numerous studies have attempted to overcome these limitations, it remains difficult to clearly distinguish the performance gains of multispectral detection systems from the impact of these "optimization techniques". Worse still, despite the rapid emergence of high-performing single-modality detection models, there is still a lack of specialized training techniques that can effectively adapt these models for multispectral detection tasks. The absence of a standardized benchmark with fair and consistent experimental setups also poses a significant barrier to evaluating the effectiveness of new approaches. To this end, we propose the first fair and reproducible benchmark specifically designed to evaluate the training "techniques", which systematically classifies existing multispectral object detection methods, investigates their sensitivity to hyper-parameters, and standardizes the core configurations. A comprehensive evaluation is conducted across multiple representative multispectral object detection datasets, utilizing various backbone networks and detection frameworks. Additionally, we introduce an efficient and easily deployable multispectral object detection framework that can seamlessly optimize high-performing single-modality models into dual-modality models, integrating our advanced training techniques.

Exotic hadrons beyond the conventional quark model have been discovered in the past two decades. Investigations of these states can lead to deep understanding of nonperturbative dynamics of the strong interaction. In this concise review, we focus on the productions of exotic hadrons in $pp$, $p\bar p$, and nuclear collisions. Experimental observations of light nuclei and hypernuclei, as prototypes of hadronic molecules, in heavy ion collisions will also be briefly discussed., Comment: 17 pages, 16 figures. This review is dedicated to Professor Wenqing Shen on the occasion to celebrate his 80th birthday. All authors contributed equally to this work

In autonomous driving, recent advances in lane segment perception provide autonomous vehicles with a comprehensive understanding of driving scenarios. Moreover, incorporating prior information input into such perception model represents an effective approach to ensure the robustness and accuracy. However, utilizing diverse sources of prior information still faces three key challenges: the acquisition of high-quality prior information, alignment between prior and online perception, efficient integration. To address these issues, we investigate prior augmentation from a novel perspective of trajectory priors. In this paper, we initially extract crowdsourcing trajectory data from Argoverse2 motion forecasting dataset and encode trajectory data into rasterized heatmap and vectorized instance tokens, then we incorporate such prior information into the online mapping model through different ways. Besides, with the purpose of mitigating the misalignment between prior and online perception, we design a confidence-based fusion module that takes alignment into account during the fusion process. We conduct extensive experiments on OpenLane-V2 dataset. The results indicate that our method's performance significantly outperforms the current state-of-the-art methods.

A recursive extension of the hybrid tetrahedron method for Brillouin-zone integration is proposed, allowing iterative tetrahedron refinement and significantly reducing the error from the linear tetrahedron method. The Brillouin-zone integral is expressed as a weighted sum on the initial grid, with integral weights collected recursively from the finest grid. Our method is capable of simultaneously handling multiple singularities in the integrand and thus may provide practical solutions to various Brillouin-zone integral tasks encountered in realistic calculations, including the computation of response and spectral function with superior sampling convergence. We demonstrate its effectiveness through numerical calculations of the density response functions of two model Hamiltonians and one real material system, the face-centered cubic cobalt., Comment: 15 pages, 6 figures

To guide a learner to master the action skills, it is crucial for a coach to 1) reason through the learner's action execution and technical keypoints, and 2) provide detailed, understandable feedback on what is done well and what can be improved. However, existing score-based action assessment methods are still far from this practical scenario. To bridge this gap, we investigate a new task termed Descriptive Action Coaching (DAC) which requires a model to provide detailed commentary on what is done well and what can be improved beyond a quality score from an action execution. To this end, we construct a new dataset named EE4D-DAC. With an LLM-based annotation pipeline, our dataset goes beyond the existing action assessment datasets by providing the hierarchical coaching commentary at both keypoint and instance levels. Furthermore, we propose TechCoach, a new framework that explicitly incorporates keypoint-level reasoning into the DAC process. The central to our method lies in the Context-aware Keypoint Reasoner, which enables TechCoach to learn keypoint-related quality representations by querying visual context under the supervision of keypoint-level coaching commentary. Prompted by the visual context and the keypoint-related quality representations, a unified Keypoint-aware Action Assessor is then employed to provide the overall coaching commentary together with the quality score. Combining all of these, we build a new benchmark for DAC and evaluate the effectiveness of our method through extensive experiments. Data and code will be publicly available., Comment: 19 pages, 12 figures

Ensuring safe, comfortable, and efficient navigation is a critical goal for autonomous driving systems. While end-to-end models trained on large-scale datasets excel in common driving scenarios, they often struggle with rare, long-tail events. Recent progress in large language models (LLMs) has introduced enhanced reasoning capabilities, but their computational demands pose challenges for real-time decision-making and precise planning. This paper presents FASIONAD, a novel dual-system framework inspired by the cognitive model "Thinking, Fast and Slow." The fast system handles routine navigation tasks using rapid, data-driven path planning, while the slow system focuses on complex reasoning and decision-making in challenging or unfamiliar situations. A dynamic switching mechanism based on score distribution and feedback allows seamless transitions between the two systems. Visual prompts generated by the fast system enable human-like reasoning in the slow system, which provides high-quality feedback to enhance the fast system's decision-making. To evaluate FASIONAD, we introduce a new benchmark derived from the nuScenes dataset, specifically designed to differentiate fast and slow scenarios. FASIONAD achieves state-of-the-art performance on this benchmark, establishing a new standard for frameworks integrating fast and slow cognitive processes in autonomous driving. This approach paves the way for more adaptive, human-like autonomous driving systems.

Remote sensing foundation models largely break away from the traditional paradigm of designing task-specific models, offering greater scalability across multiple tasks. However, they face challenges such as low computational efficiency and limited interpretability, especially when dealing with high-resolution remote sensing images. To overcome these, we draw inspiration from heat conduction, a physical process modeling local heat diffusion. Building on this idea, we are the first to explore the potential of using the parallel computing model of heat conduction to simulate the local region correlations in high-resolution remote sensing images, and introduce RS-vHeat, an efficient multi-modal remote sensing foundation model. Specifically, RS-vHeat 1) applies the Heat Conduction Operator (HCO) with a complexity of $O(N^{1.5})$ and a global receptive field, reducing computational overhead while capturing remote sensing object structure information to guide heat diffusion; 2) learns the frequency distribution representations of various scenes through a self-supervised strategy based on frequency domain hierarchical masking and multi-domain reconstruction; 3) significantly improves efficiency and performance over state-of-the-art techniques across 4 tasks and 10 datasets. Compared to attention-based remote sensing foundation models, we reduces memory consumption by 84%, decreases FLOPs by 24% and improves throughput by 2.7 times., Comment: 18 pages, 9 figures and 9 tables

In the field of finance, the prediction of individual credit default is of vital importance. However, existing methods face problems such as insufficient interpretability and transparency as well as limited performance when dealing with high-dimensional and nonlinear data. To address these issues, this paper introduces a method based on Kolmogorov-Arnold Networks (KANs). KANs is a new type of neural network architecture with learnable activation functions and no linear weights, which has potential advantages in handling complex multi-dimensional data. Specifically, this paper applies KANs to the field of individual credit risk prediction for the first time and constructs the Kolmogorov-Arnold Credit Default Predict (KACDP) model. Experiments show that the KACDP model outperforms mainstream credit default prediction models in performance metrics (ROC_AUC and F1 values). Meanwhile, through methods such as feature attribution scores and visualization of the model structure, the model's decision-making process and the importance of different features are clearly demonstrated, providing transparent and interpretable decision-making basis for financial institutions and meeting the industry's strict requirements for model interpretability. In conclusion, the KACDP model constructed in this paper exhibits excellent predictive performance and satisfactory interpretability in individual credit risk prediction, providing an effective way to address the limitations of existing methods and offering a new and practical credit risk prediction tool for financial institutions.

An object moving with the acceleration will change the temperature of environment around it, because of the presence of the Unruh thermal effect. In this work, we investigate the impact of Unruh thermal noise on the quantum-memory-assisted {entropic} uncertainty and quantum correlation regarding a pair of Unruh-Dewitt detectors. Specifically, we examine how the acceleration, the coupling strength between the external field and the detector, and the initial state affect the uncertainty and the system's quantum discord. It turns out that the Unruh effect will result in the loss of the systemic quantumness and inflation of the uncertainty. Moreover, it is revealed that the uncertainty is reversely correlated with the system's quantum discord. Thereby, it is believed that our investigations provide new insights into understanding the behavior of objects in the relativistic background., Comment: 6 pages, 5 figures, comments are welcomed. Accepted by Physics Letters B

The X-ray polarization observations made possible with the Imaging X-ray Polarimetry Explorer (IXPE) offer new ways of probing high-energy emission processes in astrophysical jets from blazars. Here we report on the first X-ray polarization observation of the blazar S4 0954+65 in a high optical and X-ray state. During our multi-wavelength campaign on the source, we detected an optical flare whose peak coincided with the peak of an X-ray flare. This optical-X-ray flare most likely took place in a feature moving along the parsec-scale jet, imaged at 43 GHz by the Very Long Baseline Array. The 43 GHz polarization angle of the moving component underwent a rotation near the time of the flare. In the optical band, prior to the IXPE observation, we measured the polarization angle to be aligned with the jet axis. In contrast, during the optical flare the optical polarization angle was perpendicular to the jet axis; after the flare, it reverted to being parallel to the jet axis. Due to the smooth behavior of the optical polarization angle during the flare, we favor shocks as the main acceleration mechanism. We also infer that the ambient magnetic field lines in the jet were parallel to the jet position angle. The average degree of optical polarization during the IXPE observation was (14.3$\pm$4.1)%. Despite the flare, we only detected an upper limit of 14% (at 3$\sigma$ level) on the X-ray polarization degree; although a reasonable assumption on the X-ray polarization angle results in an upper limit of 8.8% ($3\sigma$). We model the spectral energy distribution (SED) and spectral polarization distribution (SPD) of S4 0954+65 with leptonic (synchrotron self-Compton) and hadronic (proton and pair synchrotron) models. The constraints we obtain with our combined multi-wavelength polarization observations and SED modeling tentatively disfavor hadronic models for the X-ray emission in S4 0954+65., Comment: Submitted to A&A, 16 pages, 5 figures, and 7 tables

Self-improvement in multimodal large language models (MLLMs) is crucial for enhancing their reliability and robustness. However, current methods often rely heavily on MLLMs themselves as judges, leading to high computational costs and potential pitfalls like reward hacking and model collapse. This paper introduces a novel, model-level judge-free self-improvement framework. Our approach employs a controlled feedback mechanism while eliminating the need for MLLMs in the verification loop. We generate preference learning pairs using a controllable hallucination mechanism and optimize data quality by leveraging lightweight, contrastive language-image encoders to evaluate and reverse pairs when necessary. Evaluations across public benchmarks and our newly introduced IC dataset designed to challenge hallucination control demonstrate that our model outperforms conventional techniques. We achieve superior precision and recall with significantly lower computational demands. This method offers an efficient pathway to scalable self-improvement in MLLMs, balancing performance gains with reduced resource requirements.

Achieving optimal performance of video diffusion transformers within given data and compute budget is crucial due to their high training costs. This necessitates precisely determining the optimal model size and training hyperparameters before large-scale training. While scaling laws are employed in language models to predict performance, their existence and accurate derivation in visual generation models remain underexplored. In this paper, we systematically analyze scaling laws for video diffusion transformers and confirm their presence. Moreover, we discover that, unlike language models, video diffusion models are more sensitive to learning rate and batch size, two hyperparameters often not precisely modeled. To address this, we propose a new scaling law that predicts optimal hyperparameters for any model size and compute budget. Under these optimal settings, we achieve comparable performance and reduce inference costs by 40.1% compared to conventional scaling methods, within a compute budget of 1e10 TFlops. Furthermore, we establish a more generalized and precise relationship among validation loss, any model size, and compute budget. This enables performance prediction for non-optimal model sizes, which may also be appealed under practical inference cost constraints, achieving a better trade-off.

Superconducting magnets with non-circular apertures are important for handling unconventional beam profiles and specialized accelerator applications. This paper presents an analytical framework for designing superconducting accelerator magnets with quasi-polygonal apertures, aimed at generating precise circular field harmonics. In Part 1, we explore the relationship between current distributions on quasi-polygonal formers and their corresponding magnetic field harmonics. By employing conformal mapping techniques, we establish a connection between the design of quasi-polygonal bore magnets and traditional circular bore configurations, facilitating the simplification of complex mathematical formulations. Part 2 applies the derived current distributions to the canted cosine theta (CCT) coil magnet concept, focusing on designing analytic winding schemes that generate single or mixed circular harmonics within quasi-polygonal apertures. This work not only advances the design of superconducting magnets but also broadens the scope of CCT technology to accommodate more complex geometries., Comment: 11 pages, 29 figures

Large Language Models (LLMs) have strong instruction-following capability to interpret and execute tasks as directed by human commands. Multimodal Large Language Models (MLLMs) have inferior instruction-following ability compared to LLMs. However, there is a significant gap in the instruction-following capabilities between the MLLMs and LLMs. In this study, we conduct a pilot experiment, which demonstrates that spatially down-sampling visual tokens significantly enhances the instruction-following capability of MLLMs. This is attributed to the substantial redundancy in visual modality. However, this intuitive method severely impairs the MLLM's multimodal understanding capability. In this paper, we propose Visual-Modality Token Compression (VMTC) and Cross-Modality Attention Inhibition (CMAI) strategies to alleviate this gap between MLLMs and LLMs by inhibiting the influence of irrelevant visual tokens during content generation, increasing the instruction-following ability of the MLLMs while retaining their multimodal understanding capacity. In VMTC module, the primary tokens are retained and the redundant tokens are condensed by token clustering and merging. In CMAI process, we aggregate text-to-image attentions by text-to-text attentions to obtain a text-to-image focus score. Attention inhibition is performed on the text-image token pairs with low scores. Our comprehensive experiments over instruction-following capabilities and VQA-V2, GQA, TextVQA, MME and MMBench five benchmarks, demonstrate that proposed strategy significantly enhances the instruction following capability of MLLMs while preserving the ability to understand and process multimodal inputs.

Surgical practice involves complex visual interpretation, procedural skills, and advanced medical knowledge, making surgical vision-language pretraining (VLP) particularly challenging due to this complexity and the limited availability of annotated data. To address the gap, we propose OphCLIP, a hierarchical retrieval-augmented vision-language pretraining framework specifically designed for ophthalmic surgical workflow understanding. OphCLIP leverages the OphVL dataset we constructed, a large-scale and comprehensive collection of over 375K hierarchically structured video-text pairs with tens of thousands of different combinations of attributes (surgeries, phases/operations/actions, instruments, medications, as well as more advanced aspects like the causes of eye diseases, surgical objectives, and postoperative recovery recommendations, etc). These hierarchical video-text correspondences enable OphCLIP to learn both fine-grained and long-term visual representations by aligning short video clips with detailed narrative descriptions and full videos with structured titles, capturing intricate surgical details and high-level procedural insights, respectively. Our OphCLIP also designs a retrieval-augmented pretraining framework to leverage the underexplored large-scale silent surgical procedure videos, automatically retrieving semantically relevant content to enhance the representation learning of narrative videos. Evaluation across 11 datasets for phase recognition and multi-instrument identification shows OphCLIP's robust generalization and superior performance.

This paper studies decentralized bilevel optimization, in which multiple agents collaborate to solve problems involving nested optimization structures with neighborhood communications. Most existing literature primarily utilizes gradient tracking to mitigate the influence of data heterogeneity, without exploring other well-known heterogeneity-correction techniques such as EXTRA or Exact Diffusion. Additionally, these studies often employ identical decentralized strategies for both upper- and lower-level problems, neglecting to leverage distinct mechanisms across different levels. To address these limitations, this paper proposes SPARKLE, a unified Single-loop Primal-dual AlgoRithm frameworK for decentraLized bilEvel optimization. SPARKLE offers the flexibility to incorporate various heterogeneitycorrection strategies into the algorithm. Moreover, SPARKLE allows for different strategies to solve upper- and lower-level problems. We present a unified convergence analysis for SPARKLE, applicable to all its variants, with state-of-the-art convergence rates compared to existing decentralized bilevel algorithms. Our results further reveal that EXTRA and Exact Diffusion are more suitable for decentralized bilevel optimization, and using mixed strategies in bilevel algorithms brings more benefits than relying solely on gradient tracking., Comment: 73 pages, the Thirty-Eighth Annual Conference on Neural Information Processing Systems (2024)

This paper presents an enhanced optical configuration for a single-pass quantitative Schlieren imaging system that achieves an optical resolution of approximately 4.6 micrometers. The modified setup decouples sensitivity from resolution, enabling independent optimization of these critical parameters. Using this high-resolution system, we conduct quantitative analyses of supersonic jets emitted from sub-millimeter nozzles into the atmosphere and investigate shock waves induced by knife blades interacting with these jets in a vacuum environment. The fine resolution allows for detailed visualization of shock wave structures and accurate measurement of density gradients. We demonstrate the system's effectiveness by examining the density gradient profile along the shock diamonds and mapping density profiles across shock waves. These density profiles are analyzed for their relevance in laser-plasma applications, including laser wakefield acceleration and the Analog Black Hole Evaporation via Laser (AnaBHEL) experiment. Our findings indicate that this system can help determine key parameters such as peak density, plateau length, and shock wave thickness-essential for optimizing electron acceleration and achieving specific plasma density profiles. This high-resolution quantitative Schlieren imaging technique thus serves as a valuable tool for exploring complex fluid dynamics and supporting advancements in laser-plasma physics research., Comment: 16 pages, 13 figures

Being closely connected to the origin of the nucleon mass, the gravitational form factors of the nucleon have attracted significant attention in recent years. We present the first model-independent precise determinations of the gravitational form factors of the pion and nucleon at the physical pion mass, using a data-driven dispersive approach. The so-called ``last unknown global property'' of the nucleon, the $D$-term, is determined to be $-\left(3.38^{+0.26}_{-0.32}\right)$. The root mean square radius of the mass distribution inside the nucleon is determined to be $0.97^{+0.02}_{-0.03}~\text{fm}$. Notably, this value is larger than the proton charge radius, suggesting a modern structural view of the nucleon where gluons, responsible for most of the nucleon mass, are distributed over a larger spatial region than quarks, which dominate the charge distribution. We also predict the nucleon angular momentum and mechanical radii, providing further insights into the intricate internal structure of the nucleon., Comment: 8 pages, 6 figures

The scaling law is a notable property of neural network models and has significantly propelled the development of large language models. Scaling laws hold great promise in guiding model design and resource allocation. Recent research increasingly shows that scaling laws are not limited to NLP tasks or Transformer architectures; they also apply to domains such as recommendation. However, there is still a lack of literature on scaling law research in online advertisement retrieval systems. This may be because 1) identifying the scaling law for resource cost and online revenue is often expensive in both time and training resources for large-scale industrial applications, and 2) varying settings for different systems prevent the scaling law from being applied across various scenarios. To address these issues, we propose a lightweight paradigm to identify the scaling law of online revenue and machine cost for a certain online advertisement retrieval scenario with a low experimental cost. Specifically, we focus on a sole factor (FLOPs) and propose an offline metric named R/R* that exhibits a high linear correlation with online revenue for retrieval models. We estimate the machine cost offline via a simulation algorithm. Thus, we can transform most online experiments into low-cost offline experiments. We conduct comprehensive experiments to verify the effectiveness of our proposed metric R/R* and to identify the scaling law in the online advertisement retrieval system of Kuaishou. With the scaling law, we demonstrate practical applications for ROI-constrained model designing and multi-scenario resource allocation in Kuaishou advertising system. To the best of our knowledge, this is the first work to study the scaling laws for online advertisement retrieval of real-world systems, showing great potential for scaling law in advertising system optimization., Comment: 10 pages, 8 figures

Recently, Large Language Models (LLMs)-based multi-agent paradigms for software engineering are introduced to automatically resolve software development tasks (e.g., from a given issue to source code). However, existing work is evaluated based on historical benchmark datasets, does not consider human feedback at each stage of the automated software development process, and has not been deployed in practice. In this paper, we introduce a Human-in-the-loop LLM-based Agents framework (HULA) for software development that allows software engineers to refine and guide LLMs when generating coding plans and source code for a given task. We design, implement, and deploy the HULA framework into Atlassian JIRA for internal uses. Through a multi-stage evaluation of the HULA framework, Atlassian software engineers perceive that HULA can minimize the overall development time and effort, especially in initiating a coding plan and writing code for straightforward tasks. On the other hand, challenges around code quality are raised to be solved in some cases. We draw lessons learned and discuss opportunities for future work, which will pave the way for the advancement of LLM-based agents in software development.

In this work, we present a novel method called the complex frequency fingerprint (CFF) to detect the complex frequency Green's function, $G(\omega\in\mathbb{C})$, in a driven-dissipative system. By utilizing the CFF, we can measure the complex frequency density of states (DOS) and local DOS (LDOS), providing unique insights into the characterization of non-Hermitian systems. By applying our method to systems exhibiting the non-Hermitian skin effect (NHSE), we demonstrate how to use our theory to detect both the non-Hermitian eigenvalues and eigenstates. This offers a distinctive and reliable approach to identifying the presence or absence of NHSE in experimental settings., Comment: 4+5 pages, 4+1 figures, comments welcome!

Quantum computing solutions are increasingly deployed in commercial environments through delegated computing, especially one of the most critical issues is to guarantee the confidentiality and proprietary of quantum implementations. Since the proposal of general-purpose indistinguishability obfuscation (iO) and functional encryption schemes, iO has emerged as a seemingly versatile cryptography primitive. Existing research on quantum indistinguishable obfuscation (QiO) primarily focuses on task-oriented, lacking solutions to general quantum computing. In this paper, we propose a scheme for constructing QiO via the equivalence of quantum circuits. It introduces the concept of quantum subpath sum equivalence, demonstrating that indistinguishability between two quantum circuits can be achieved by incremental changes in quantum subpaths. The restriction of security loss is solved by reducing the distinguisher to polynomial probability test. The scheme obfuscates the quantum implementation of classical functions in a path-sum specification, ensuring the indistinguishability between different quantum implementations. The results demonstrate the feasibility of indistinguishability obfuscation for general circuits and provide novel insights on intellectual property protection and secure delegated quantum computing.

This paper investigates the initial boundary value problem of finitely degenerate semilinear pseudo-parabolic equations associated with H\"{o}rmander's operator. For the low and critical initial energies, based on the global existence of solutions in the previous literature, we investigate the exponential decay estimate of the energy functional. Moreover, we provide a method different from the previous literature to prove the finite time blow-up of solutions under low and critical initial energies, and derive the upper bounds of blow-up time and the rate as well as the exponential growth estimate of blow-up solutions. Furthermore, we prove the existence of global solutions and blow-up solutions at high initial energy. In particular, we establish a blow-up condition independent of the mountain pass level and provide the upper bound of blow-up time and the exponential growth estimate of blow-up solutions, and further we prove the existence of blow-up solutions with arbitrary initial energy by utilizing this blow-up condition. Finally, we determine a lower bound of the blow-up rate for all the above-mentioned blow-up solutions. These results complement the ones obtained in the previous literature.

In this paper, we address the challenging task of multimodal mathematical reasoning by incorporating the ability of ``slow thinking" into multimodal large language models (MLLMs). Contrary to existing methods that rely on direct or fast thinking, our key idea is to construct long chains of thought (CoT) consisting of atomic actions in a step-by-step manner, guiding MLLMs to perform complex reasoning. To this end, we design a novel AtomThink framework composed of three key modules: (i) a CoT annotation engine that automatically generates high-quality CoT annotations to address the lack of high-quality visual mathematical data; (ii) an atomic step fine-tuning strategy that jointly optimizes an MLLM and a policy reward model (PRM) for step-wise reasoning; and (iii) four different search strategies that can be applied with the PRM to complete reasoning. Additionally, we propose AtomMATH, a large-scale multimodal dataset of long CoTs, and an atomic capability evaluation metric for mathematical tasks. Extensive experimental results show that the proposed AtomThink significantly improves the performance of baseline MLLMs, achieving approximately 50\% relative accuracy gains on MathVista and 120\% on MathVerse. To support the advancement of multimodal slow-thinking models, we will make our code and dataset publicly available on https://github.com/Quinn777/AtomThink.

Continuous-valued Autoregressive (AR) image generation models have demonstrated notable superiority over their discrete-token counterparts, showcasing considerable reconstruction quality and higher generation fidelity. However, the computational demands of the autoregressive framework result in significant inference overhead. While speculative decoding has proven effective in accelerating Large Language Models (LLMs), their adaptation to continuous-valued visual autoregressive models remains unexplored. This work generalizes the speculative decoding algorithm from discrete tokens to continuous space. By analyzing the intrinsic properties of output distribution, we establish a tailored acceptance criterion for the diffusion distributions prevalent in such models. To overcome the inconsistency that occurred in speculative decoding output distributions, we introduce denoising trajectory alignment and token pre-filling methods. Additionally, we identify the hard-to-sample distribution in the rejection phase. To mitigate this issue, we propose a meticulous acceptance-rejection sampling method with a proper upper bound, thereby circumventing complex integration. Experimental results show that our continuous speculative decoding achieves a remarkable $2.33\times$ speed-up on off-the-shelf models while maintaining the output distribution. Codes will be available at https://github.com/MarkXCloud/CSpD

Efficient acquisition of real-world embodied data has been increasingly critical. However, large-scale demonstrations captured by remote operation tend to take extremely high costs and fail to scale up the data size in an efficient manner. Sampling the episodes under a simulated environment is a promising way for large-scale collection while existing simulators fail to high-fidelity modeling on texture and physics. To address these limitations, we introduce the RoboGSim, a real2sim2real robotic simulator, powered by 3D Gaussian Splatting and the physics engine. RoboGSim mainly includes four parts: Gaussian Reconstructor, Digital Twins Builder, Scene Composer, and Interactive Engine. It can synthesize the simulated data with novel views, objects, trajectories, and scenes. RoboGSim also provides an online, reproducible, and safe evaluation for different manipulation policies. The real2sim and sim2real transfer experiments show a high consistency in the texture and physics. Moreover, the effectiveness of synthetic data is validated under the real-world manipulated tasks. We hope RoboGSim serves as a closed-loop simulator for fair comparison on policy learning. More information can be found on our project page https://robogsim.github.io/ .

Sapphire is an attractive material in photonic, optoelectronic, and transparent ceramic applications that stand to benefit from surface functionalization effects stemming from micro/nanostructures. Here we investigate the use of ultrafast lasers for fabricating nanostructures in sapphire by exploring the relationship between irradiation parameters, morphology change, and selective etching. In this approach an ultrafast laser pulse is focused on the sapphire substrate to change the crystalline morphology to amorphous or polycrystalline, which is characterized by examining different vibrational modes using Raman spectroscopy. The irradiated regions are then removed using a subsequent wet etch in hydrofluoric acid. Laser confocal measurements conducted before and after the etching process quantify the degree of selective etching. The results indicate that a threshold laser pulse intensity is required for selective etching to occur. This process can be used to fabricate hierarchical sapphire nanostructures over large areas with enhanced hydrophobicity, which exhibits an apparent contact angle of 140 degrees and a high roll-off angle that are characteristic of the rose petal effect. Additionally, the fabricated structures have high broadband diffuse transmittance of up to 81.8% with low loss, which can find applications in optical diffusers. Our findings provide new insights into the interplay between the light-matter interactions, where Raman shifts associated with different vibrational modes can be used as a predictive measure of selective etching. These results advance the development of sapphire nanostructure fabrication, which can find applications in infrared optics, protective windows, and consumer electronics., Comment: 41 pages, 21 figures

In recent years, with the significant evolution of multi-modal large models, many recommender researchers realized the potential of multi-modal information for user interest modeling. In industry, a wide-used modeling architecture is a cascading paradigm: (1) first pre-training a multi-modal model to provide omnipotent representations for downstream services; (2) The downstream recommendation model takes the multi-modal representation as additional input to fit real user-item behaviours. Although such paradigm achieves remarkable improvements, however, there still exist two problems that limit model performance: (1) Representation Unmatching: The pre-trained multi-modal model is always supervised by the classic NLP/CV tasks, while the recommendation models are supervised by real user-item interaction. As a result, the two fundamentally different tasks' goals were relatively separate, and there was a lack of consistent objective on their representations; (2) Representation Unlearning: The generated multi-modal representations are always stored in cache store and serve as extra fixed input of recommendation model, thus could not be updated by recommendation model gradient, further unfriendly for downstream training. Inspired by the two difficulties challenges in downstream tasks usage, we introduce a quantitative multi-modal framework to customize the specialized and trainable multi-modal information for different downstream models., Comment: Work in progress

In the previous article, we introduced a neural network framework based on symmetric differential equations. This novel framework exhibits complete symmetry, endowing it with perfect mathematical properties. While we have examined some of the system's mathematical characteristics, a detailed discussion of the network training methodology has not yet been presented. Drawing on the principles of the traditional backpropagation algorithm, this study proposes an alternative training approach that utilizes differential equation signal propagation instead of chain rule derivation. This approach not only preserves the effectiveness of training but also offers enhanced biological interpretability. The foundation of this methodology lies in the system's reversibility, which stems from its inherent symmetry,a key aspect of our research. However, this method alone is insufficient for effective neural network training. To address this, we further introduce a distributed Proportional-Integral-Derivative (PID) control approach, emphasizing its implementation within a closed system. By incorporating this method, we achieved both faster training speeds and improved accuracy. This approach not only offers novel insights into neural network training but also extends the scope of research into control methodologies. To validate its effectiveness, we apply this method to the MNIST dataset, demonstrating its practical utility., Comment: 12 pages, 5 figures

A new caesium (Cs) fountain clock NIM6 has been developed at the National Institute of Metrology (NIM) in China, for which a comprehensive uncertainty evaluation is presented. A three-dimensional magneto-optical trap (3D MOT) loading optical molasses is employed to obtain more cold atoms rapidly and efficiently with a tunable, uniform density distribution. A heat pipe surrounding the flight tube maintains a consistent and stable temperature within the interrogation region. Additionally, a Ramsey cavity with four azimuthally distribution feeds is utilized to mitigate distributed cavity phase shifts. The Cs fountain clock NIM6 achieves a short-term stability of 1.0x10-13 {\tau}-1/2 at high atomic density, and a typical overall fractional type-B uncertainty is estimated to be 2.3x10-16. Comparisons of frequency between the Cs fountain NIM6 and other Cs fountain Primary Frequency Standards (PFSs) through Coordinated Universal Time (UTC) have demonstrated an agreement within the stated uncertainties., Comment: 14 pages, 16 figures

In a recent paper, Phys. Rev. Lett. {\bf 126}, 132001 (2021), the LHCb data on the di-$J/\psi$ production in proton-proton collisions were analysed in a coupled-channel framework based on double-vector-charmonium channels. This investigation identified a robust pole near the $J/\psi J/\psi$ threshold, tagged $X(6200)$, suggesting it as a new state. The present work extends that investigation by incorporating recent di-$J/\psi$ production data from the CMS and ATLAS collaborations and performing a combined analysis of all three data sets. This study confirms the existence of the $X(6200)$, and its pole position is now determined with a higher precision than in the previous study, where only a single data set was employed. The pole corresponding to the $X(6900)$ is also extracted, though its actual position and residue depend on a particular coupled-channel model employed and deviate from values reported in the experimental investigations. The dielectron width of the $X(6200)$ is estimated under different conjectures on its quantum numbers, and the number of such states that can be annually produced at the future Super $\tau$-Charm Facility is estimated. The line shape in the complimentary $J/\psi\psi(2S)$ channel is discussed, and a good agreement with the ATLAS data is found., Comment: 9 pages, 5 figures, 7 tables

By applying some techniques of set-valued and variational analysis, we study solution stability of nonhomogeneous split equality problems and nonhomogeneous split feasibility problems, where the constraint sets need not be convex. Necessary and sufficient conditions for the Lipschitz-likeness of the solution maps of the problems are given and illustrated by concrete examples. The obtained results complement those given in [Huong VT, Xu HK, Yen ND. Stability analysis of split equality and split feasibility problems. arXiv:2410.16856.], where classical split equality problems and split feasibility problems have been considered.

In Radio Super Novae (RSNe) a magnetic field of $(B \, \times \, r) \, = \, 10^{16.0 \pm 0.12} \, {\rm Gauss \, \times \, cm}$ is observed; these are the same numbers for Blue Super Giant (BSG) star explosions as for Red Super Giant (RSG) star explosions, despite their very different wind properties. The EHT data for M87 as well for low power radio galaxies all show consistency with just this value of the quantity $(B \, \times \, r )$, key for angular momentum and energy transport, and can be derived from the radio jet data. We interpret this as a property of the near surroundings of a black hole (BH) at near maximal rotation, independent of BH mass. In the commonly used green onion model, in which a $2 \, \pi$ flow changes over to a jet flow we interpret this as a wind emanating from the BH/accretion disk system and its surroundings. Near the BH collisions in the wind can produce a large fraction of anti-protons. In this scenario the cosmic Ray (CR) population from the wind/jet is proposed to be visible as EeV protons and anti-protons in the CR data to EeV energy, with a $E^{-7/3}$ spectrum. This can be connected to a concept of inner and outer Penrose zones in the ergo-region. The observed numbers for the magnetic field imply the Planck time as the governing time scale: A BH rotating near maximum can accept a proton per log bin of energy in an extended spectrum with the associated pions every Planck time., Comment: 81 pages, 5 figures, invited paper (in press), Frontiers in Astronomy and Space Sciences

Autonomous driving evaluation requires simulation environments that closely replicate actual road conditions, including real-world sensory data and responsive feedback loops. However, many existing simulations need to predict waypoints along fixed routes on public datasets or synthetic photorealistic data, \ie, open-loop simulation usually lacks the ability to assess dynamic decision-making. While the recent efforts of closed-loop simulation offer feedback-driven environments, they cannot process visual sensor inputs or produce outputs that differ from real-world data. To address these challenges, we propose DrivingSphere, a realistic and closed-loop simulation framework. Its core idea is to build 4D world representation and generate real-life and controllable driving scenarios. In specific, our framework includes a Dynamic Environment Composition module that constructs a detailed 4D driving world with a format of occupancy equipping with static backgrounds and dynamic objects, and a Visual Scene Synthesis module that transforms this data into high-fidelity, multi-view video outputs, ensuring spatial and temporal consistency. By providing a dynamic and realistic simulation environment, DrivingSphere enables comprehensive testing and validation of autonomous driving algorithms, ultimately advancing the development of more reliable autonomous cars. The benchmark will be publicly released., Comment: https://yanty123.github.io/DrivingSphere/