Search

Satellite networks are able to collect massive space information with advanced remote sensing technologies, which is essential for real-time applications such as natural disaster monitoring. However, traditional centralized processing by the ground server incurs a severe timeliness issue caused by the transmission bottleneck of raw data. To this end, Space Computing Power Networks (Space-CPN) emerges as a promising architecture to coordinate the computing capability of satellites and enable on board data processing. Nevertheless, due to the natural limitations of solar panels, satellite power system is difficult to meet the energy requirements for ever-increasing intelligent computation tasks of artificial neural networks. To tackle this issue, we propose to employ spiking neural networks (SNNs), which is supported by the neuromorphic computing architecture, for on-board data processing. The extreme sparsity in its computation enables a high energy efficiency. Furthermore, to achieve effective training of these on-board models, we put forward a decentralized neuromorphic learning framework, where a communication-efficient inter-plane model aggregation method is developed with the inspiration from RelaySum. We provide a theoretical analysis to characterize the convergence behavior of the proposed algorithm, which reveals a network diameter related convergence speed. We then formulate a minimum diameter spanning tree problem on the inter-plane connectivity topology and solve it to further improve the learning performance. Extensive experiments are conducted to evaluate the superiority of the proposed method over benchmarks.

Multimodal large language models (MLLMs) have shown impressive capabilities across various domains, excelling in processing and understanding information from multiple modalities. Despite the rapid progress made previously, insufficient OCR ability hinders MLLMs from excelling in text-related tasks. In this paper, we present \textbf{Ocean-OCR}, a 3B MLLM with state-of-the-art performance on various OCR scenarios and comparable understanding ability on general tasks. We employ Native Resolution ViT to enable variable resolution input and utilize a substantial collection of high-quality OCR datasets to enhance the model performance. We demonstrate the superiority of Ocean-OCR through comprehensive experiments on open-source OCR benchmarks and across various OCR scenarios. These scenarios encompass document understanding, scene text recognition, and handwritten recognition, highlighting the robust OCR capabilities of Ocean-OCR. Note that Ocean-OCR is the first MLLM to outperform professional OCR models such as TextIn and PaddleOCR.

In vitro fertilization-embryo transfer (IVF-ET) stands as one of the most prevalent treatments for infertility. During an IVF-ET cycle, the time interval between trigger shot and oocyte pickup (OPU) is a pivotal period for follicular maturation, which determines mature oocytes yields and impacts the success of subsequent procedures. However, accurately predicting this interval is severely hindered by the variability of clinicians'experience that often leads to suboptimal oocyte retrieval rate. To address this challenge, we propose ILETIA, the first machine learning-based method that could predict the optimal trigger-OPU interval for patients receiving progestin-primed ovarian stimulation (PPOS) protocol. Specifically, ILETIA leverages a Transformer to learn representations from clinical tabular data, and then employs gradient-boosted trees for interval prediction. For model training and evaluating, we compiled a dataset PPOS-DS of nearly ten thousand patients receiving PPOS protocol, the largest such dataset to our knowledge. Experimental results demonstrate that our method achieves strong performance (AUROC = 0.889), outperforming both clinicians and other widely used computational models. Moreover, ILETIA also supports premature ovulation risk prediction in a specific OPU time (AUROC = 0.838). Collectively, by enabling more precise and individualized decisions, ILETIA has the potential to improve clinical outcomes and lay the foundation for future IVF-ET research.

We introduce Baichuan-Omni-1.5, an omni-modal model that not only has omni-modal understanding capabilities but also provides end-to-end audio generation capabilities. To achieve fluent and high-quality interaction across modalities without compromising the capabilities of any modality, we prioritized optimizing three key aspects. First, we establish a comprehensive data cleaning and synthesis pipeline for multimodal data, obtaining about 500B high-quality data (text, audio, and vision). Second, an audio-tokenizer (Baichuan-Audio-Tokenizer) has been designed to capture both semantic and acoustic information from audio, enabling seamless integration and enhanced compatibility with MLLM. Lastly, we designed a multi-stage training strategy that progressively integrates multimodal alignment and multitask fine-tuning, ensuring effective synergy across all modalities. Baichuan-Omni-1.5 leads contemporary models (including GPT4o-mini and MiniCPM-o 2.6) in terms of comprehensive omni-modal capabilities. Notably, it achieves results comparable to leading models such as Qwen2-VL-72B across various multimodal medical benchmarks.

Low-Rank Adaptation (LoRA) is widely used for adapting large language models (LLMs) to specific domains due to its efficiency and modularity. Meanwhile, vanilla LoRA struggles with task conflicts in multi-task scenarios. Recent works adopt Mixture of Experts (MoE) by treating each LoRA module as an expert, thereby mitigating task interference through multiple specialized LoRA modules. While effective, these methods often isolate knowledge within individual tasks, failing to fully exploit the shared knowledge across related tasks. In this paper, we establish a connection between single LoRA and multi-LoRA MoE, integrating them into a unified framework. We demonstrate that the dynamic routing of multiple LoRAs is functionally equivalent to rank partitioning and block-level activation within a single LoRA. We further empirically demonstrate that finer-grained LoRA partitioning, within the same total and activated parameter constraints, leads to better performance gains across heterogeneous tasks. Building on these findings, we propose Single-ranked Mixture of Experts LoRA (\textbf{SMoRA}), which embeds MoE into LoRA by \textit{treating each rank as an independent expert}. With a \textit{dynamic rank-wise activation} mechanism, SMoRA promotes finer-grained knowledge sharing while mitigating task conflicts. Experiments demonstrate that SMoRA activates fewer parameters yet achieves better performance in multi-task scenarios.

Achieving Artificial General Intelligence (AGI) requires AI agents that can not only make stratigic decisions but also engage in flexible and meaningful communication. Inspired by Wittgenstein's language game theory in Philosophical Investigations, we propose that language agents can learn through in-context interaction rather than traditional multi-stage frameworks that separate decision-making from language expression. Using Werewolf, a social deduction game that tests language understanding, strategic interaction, and adaptability, we develop the Multi-agent Kahneman & Tversky's Optimization (MaKTO). MaKTO engages diverse models in extensive gameplay to generate unpaired desirable and unacceptable responses, then employs KTO to refine the model's decision-making process. In 9-player Werewolf games, MaKTO achieves a 61% average win rate across various models, outperforming GPT-4o and two-stage RL agents by relative improvements of 23.0% and 10.9%, respectively. Notably, MaKTO also demonstrates human-like performance, winning 60% against expert players and showing only 49% detectability in Turing-style blind tests. These results showcase MaKTO's superior decision-making, strategic adaptation, and natural language generation in complex social deduction games., Comment: Preprint. Code and data will be available at https://reneeye.github.io/MaKTO.html

Existing zero-shot temporal action detection (ZSTAD) methods predominantly use fully supervised or unsupervised strategies to recognize unseen activities. However, these training-based methods are prone to domain shifts and require high computational costs, which hinder their practical applicability in real-world scenarios. In this paper, unlike previous works, we propose a training-Free Zero-shot temporal Action Detection (FreeZAD) method, leveraging existing vision-language (ViL) models to directly classify and localize unseen activities within untrimmed videos without any additional fine-tuning or adaptation. We mitigate the need for explicit temporal modeling and reliance on pseudo-label quality by designing the LOGarithmic decay weighted Outer-Inner-Contrastive Score (LogOIC) and frequency-based Actionness Calibration. Furthermore, we introduce a test-time adaptation (TTA) strategy using Prototype-Centric Sampling (PCS) to expand FreeZAD, enabling ViL models to adapt more effectively for ZSTAD. Extensive experiments on the THUMOS14 and ActivityNet-1.3 datasets demonstrate that our training-free method outperforms state-of-the-art unsupervised methods while requiring only 1/13 of the runtime. When equipped with TTA, the enhanced method further narrows the gap with fully supervised methods.

We introduce Sigma, an efficient large language model specialized for the system domain, empowered by a novel architecture including DiffQKV attention, and pre-trained on our meticulously collected system domain data. DiffQKV attention significantly enhances the inference efficiency of Sigma by optimizing the Query (Q), Key (K), and Value (V) components in the attention mechanism differentially, based on their varying impacts on the model performance and efficiency indicators. Specifically, we (1) conduct extensive experiments that demonstrate the model's varying sensitivity to the compression of K and V components, leading to the development of differentially compressed KV, and (2) propose augmented Q to expand the Q head dimension, which enhances the model's representation capacity with minimal impacts on the inference speed. Rigorous theoretical and empirical analyses reveal that DiffQKV attention significantly enhances efficiency, achieving up to a 33.36% improvement in inference speed over the conventional grouped-query attention (GQA) in long-context scenarios. We pre-train Sigma on 6T tokens from various sources, including 19.5B system domain data that we carefully collect and 1T tokens of synthesized and rewritten data. In general domains, Sigma achieves comparable performance to other state-of-arts models. In the system domain, we introduce the first comprehensive benchmark AIMicius, where Sigma demonstrates remarkable performance across all tasks, significantly outperforming GPT-4 with an absolute improvement up to 52.5%.

To address the high risks associated with improper use of safety gear in complex power line environments, where target occlusion and large variance are prevalent, this paper proposes an enhanced PEC-YOLO object detection algorithm. The method integrates deep perception with multi-scale feature fusion, utilizing PConv and EMA attention mechanisms to enhance feature extraction efficiency and minimize model complexity. The CPCA attention mechanism is incorporated into the SPPF module, improving the model's ability to focus on critical information and enhance detection accuracy, particularly in challenging conditions. Furthermore, the introduction of the BiFPN neck architecture optimizes the utilization of low-level and high-level features, enhancing feature representation through adaptive fusion and context-aware mechanism. Experimental results demonstrate that the proposed PEC-YOLO achieves a 2.7% improvement in detection accuracy compared to YOLOv8s, while reducing model parameters by 42.58%. Under identical conditions, PEC-YOLO outperforms other models in detection speed, meeting the stringent accuracy requirements for safety gear detection in construction sites. This study contributes to the development of efficient and accurate intelligent monitoring systems for ensuring worker safety in hazardous environments.

This paper introduces an innovative drone carrier concept that is applied in maritime port security or offshore rescue. This system works with a heterogeneous system consisting of multiple Unmanned Aerial Vehicles (UAVs) and Unmanned Surface Vehicles (USVs) to perform inspection and intervention tasks in GNSS-denied or interrupted environments. The carrier, an electric catamaran measuring 4m by 7m, features a 4m by 6m deck supporting automated takeoff and landing for four DJI M300 drones, along with a 10kg-payload manipulator operable in up to level 3 sea conditions. Utilizing an offshore gimbal camera for navigation, the carrier can autonomously navigate, approach and dock with non-cooperative vessels, guided by an onboard camera, LiDAR, and Doppler Velocity Log (DVL) over a 3 km$^2$ area. UAVs equipped with onboard Ultra-Wideband (UWB) technology execute mapping, detection, and manipulation tasks using a versatile gripper designed for wet, saline conditions. Additionally, two UAVs can coordinate to transport large objects to the manipulator or interact directly with them. These procedures are fully automated and were successfully demonstrated at the Mohammed Bin Zayed International Robotic Competition (MBZIRC2024), where the drone carrier equipped with four UAVS and one manipulator, automatically accomplished the intervention tasks in sea-level-3 (wave height 1.25m) based on the rough target information., Comment: 15 pages, 12pages

We present Hunyuan3D 2.0, an advanced large-scale 3D synthesis system for generating high-resolution textured 3D assets. This system includes two foundation components: a large-scale shape generation model -- Hunyuan3D-DiT, and a large-scale texture synthesis model -- Hunyuan3D-Paint. The shape generative model, built on a scalable flow-based diffusion transformer, aims to create geometry that properly aligns with a given condition image, laying a solid foundation for downstream applications. The texture synthesis model, benefiting from strong geometric and diffusion priors, produces high-resolution and vibrant texture maps for either generated or hand-crafted meshes. Furthermore, we build Hunyuan3D-Studio -- a versatile, user-friendly production platform that simplifies the re-creation process of 3D assets. It allows both professional and amateur users to manipulate or even animate their meshes efficiently. We systematically evaluate our models, showing that Hunyuan3D 2.0 outperforms previous state-of-the-art models, including the open-source models and closed-source models in geometry details, condition alignment, texture quality, and etc. Hunyuan3D 2.0 is publicly released in order to fill the gaps in the open-source 3D community for large-scale foundation generative models. The code and pre-trained weights of our models are available at: https://github.com/Tencent/Hunyuan3D-2, Comment: GitHub link: https://github.com/Tencent/Hunyuan3D-2

We introduce X-Dyna, a novel zero-shot, diffusion-based pipeline for animating a single human image using facial expressions and body movements derived from a driving video, that generates realistic, context-aware dynamics for both the subject and the surrounding environment. Building on prior approaches centered on human pose control, X-Dyna addresses key shortcomings causing the loss of dynamic details, enhancing the lifelike qualities of human video animations. At the core of our approach is the Dynamics-Adapter, a lightweight module that effectively integrates reference appearance context into the spatial attentions of the diffusion backbone while preserving the capacity of motion modules in synthesizing fluid and intricate dynamic details. Beyond body pose control, we connect a local control module with our model to capture identity-disentangled facial expressions, facilitating accurate expression transfer for enhanced realism in animated scenes. Together, these components form a unified framework capable of learning physical human motion and natural scene dynamics from a diverse blend of human and scene videos. Comprehensive qualitative and quantitative evaluations demonstrate that X-Dyna outperforms state-of-the-art methods, creating highly lifelike and expressive animations. The code is available at https://github.com/bytedance/X-Dyna., Comment: Project page:https://x-dyna.github.io/xdyna.github.io/ Code:https://github.com/bytedance/X-Dyna Model:https://huggingface.co/Boese0601/X-Dyna

We explore an evolutionary search strategy for scaling inference time compute in Large Language Models. The proposed approach, Mind Evolution, uses a language model to generate, recombine and refine candidate responses. The proposed approach avoids the need to formalize the underlying inference problem whenever a solution evaluator is available. Controlling for inference cost, we find that Mind Evolution significantly outperforms other inference strategies such as Best-of-N and Sequential Revision in natural language planning tasks. In the TravelPlanner and Natural Plan benchmarks, Mind Evolution solves more than 98% of the problem instances using Gemini 1.5 Pro without the use of a formal solver.

Depth completion is a key task in autonomous driving, aiming to complete sparse LiDAR depth measurements into high-quality dense depth maps through image guidance. However, existing methods usually treat depth maps as an additional channel of color images, or directly perform convolution on sparse data, failing to fully exploit the 3D geometric information in depth maps, especially with limited performance in complex boundaries and sparse areas. To address these issues, this paper proposes a depth completion network combining channel attention mechanism and 3D global feature perception (CGA-Net). The main innovations include: 1) Utilizing PointNet++ to extract global 3D geometric features from sparse depth maps, enhancing the scene perception ability of low-line LiDAR data; 2) Designing a channel-attention-based multimodal feature fusion module to efficiently integrate sparse depth, RGB images, and 3D geometric features; 3) Combining residual learning with CSPN++ to optimize the depth refinement stage, further improving the completion quality in edge areas and complex scenes. Experiments on the KITTI depth completion dataset show that CGA-Net can significantly improve the prediction accuracy of dense depth maps, achieving a new state-of-the-art (SOTA), and demonstrating strong robustness to sparse and complex scenes., Comment: 13pages,4 figures, 2 tables

Data-driven convective parameterization aims to accurately represent convective adjustments to large-scale forcings in a computationally economic manner. While previous studies have demonstrated success using various model architectures, challenges persist in developing physically interpretable models and assessing generalizability and confidence level. In this study, we develop a recurrent neural network to predict time series of temperature, moisture, and precipitation of a cumulus ensemble in response to large-scale forcings. The recurrent cell combines a linear component, pre-identified as a time-invariant state-space model within the linear limit of the problem, and a multilayer neural network for the nonlinear component. Trained on ensembles of limited-domain cloud-resolving model simulation data, the model exhibits stable and realistic performance in long-term emulations, both with prescribed large-scale forcings and when coupled with two-dimensional gravity waves. We further calculate linear responses to perturbations for the coupled emulation, revealing physically interpretable, state-dependent properties of the convectively coupled gravity wave system., Comment: 15 pages, 4 figures

Pre-trained language models (PLMs) are trained on data that inherently contains gender biases, leading to undesirable impacts. Traditional debiasing methods often rely on external corpora, which may lack quality, diversity, or demographic balance, affecting the effectiveness of debiasing. With the rise of large language models and their extensive knowledge, we propose enhancing fairness (Fair-Gender) in PLMs by absorbing coherent, attribute-balanced, and semantically rich sentences. However, these sentences cannot be directly used for debiasing due to alignment issues and the risk of negative transfer. We address this by applying causal analysis to estimate causal effects, filtering out unaligned sentences, and identifying aligned ones for incorporation into PLMs, thereby ensuring positive transfer. Experiments show that our approach significantly reduces gender biases in PLMs while preserving their language expressiveness.

Hallucination, or the generation of incorrect or fabricated information, remains a critical challenge in large language models (LLMs), particularly in high-stake domains such as legal question answering (QA). In order to mitigate the hallucination rate in legal QA, we first introduce a benchmark called LegalHalBench and three automatic metrics to evaluate the common hallucinations when LLMs answer legal questions. We then propose a hallucination mitigation method that integrates behavior cloning and a novel Hard Sample-aware Iterative Direct Preference Optimization (HIPO). We conduct extensive real-data experiments to validate the effectiveness of our approach. Our results demonstrate remarkable improvements in various metrics, including the newly proposed Non-Hallucinated Statute Rate, Statute Relevance Rate, Legal Claim Truthfulness, as well as traditional metrics such as METEOR, BERTScore, ROUGE-L, and win rates., Comment: 18 pages, 8 figures, to be published in COLING 2025

Semi-Supervised Learning (SSL) can leverage abundant unlabeled data to boost model performance. However, the class-imbalanced data distribution in real-world scenarios poses great challenges to SSL, resulting in performance degradation. Existing class-imbalanced semi-supervised learning (CISSL) methods mainly focus on rebalancing datasets but ignore the potential of using hard examples to enhance performance, making it difficult to fully harness the power of unlabeled data even with sophisticated algorithms. To address this issue, we propose a method that enhances the performance of Imbalanced Semi-Supervised Learning by Mining Hard Examples (SeMi). This method distinguishes the entropy differences among logits of hard and easy examples, thereby identifying hard examples and increasing the utility of unlabeled data, better addressing the imbalance problem in CISSL. In addition, we maintain a class-balanced memory bank with confidence decay for storing high-confidence embeddings to enhance the pseudo-labels' reliability. Although our method is simple, it is effective and seamlessly integrates with existing approaches. We perform comprehensive experiments on standard CISSL benchmarks and experimentally demonstrate that our proposed SeMi outperforms existing state-of-the-art methods on multiple benchmarks, especially in reversed scenarios, where our best result shows approximately a 54.8\% improvement over the baseline methods., Comment: 11 pages,6 figures, conference

Deep neural networks have become foundational to advancements in multiple domains, including recommendation systems, natural language processing, and so on. Despite their successes, these models often contain incompatible parameters that can be underutilized or detrimental to model performance, particularly when faced with specific, varying data distributions. Existing research excels in removing such parameters or merging the outputs of multiple different pretrained models. However, the former focuses on efficiency rather than performance, while the latter requires several times more computing and storage resources to support inference. In this paper, we set the goal to explicitly improve these incompatible parameters by leveraging the complementary strengths of different models, thereby directly enhancing the models without any additional parameters. Specifically, we propose Compatibility-aware Knowledge Integration (CKI), which consists of Parameter Compatibility Assessment and Parameter Splicing, which are used to evaluate the knowledge content of multiple models and integrate the knowledge into one model, respectively. The integrated model can be used directly for inference or for further fine-tuning. We conduct extensive experiments on various datasets for recommendation and language tasks, and the results show that Compatibility-aware Knowledge Integration can effectively optimize incompatible parameters under multiple tasks and settings to break through the training limit of the original model without increasing the inference cost., Comment: Published on AAAI'25: The Annual AAAI Conference on Artificial Intelligence

Large Language Models (LLMs) for Recommendation (LLM4Rec) is a promising research direction that has demonstrated exceptional performance in this field. However, its inability to capture real-time user preferences greatly limits the practical application of LLM4Rec because (i) LLMs are costly to train and infer frequently, and (ii) LLMs struggle to access real-time data (its large number of parameters poses an obstacle to deployment on devices). Fortunately, small recommendation models (SRMs) can effectively supplement these shortcomings of LLM4Rec diagrams by consuming minimal resources for frequent training and inference, and by conveniently accessing real-time data on devices. In light of this, we designed the Device-Cloud LLM-SRM Collaborative Recommendation Framework (LSC4Rec) under a device-cloud collaboration setting. LSC4Rec aims to integrate the advantages of both LLMs and SRMs, as well as the benefits of cloud and edge computing, achieving a complementary synergy. We enhance the practicability of LSC4Rec by designing three strategies: collaborative training, collaborative inference, and intelligent request. During training, LLM generates candidate lists to enhance the ranking ability of SRM in collaborative scenarios and enables SRM to update adaptively to capture real-time user interests. During inference, LLM and SRM are deployed on the cloud and on the device, respectively. LLM generates candidate lists and initial ranking results based on user behavior, and SRM get reranking results based on the candidate list, with final results integrating both LLM's and SRM's scores. The device determines whether a new candidate list is needed by comparing the consistency of the LLM's and SRM's sorted lists. Our comprehensive and extensive experimental analysis validates the effectiveness of each strategy in LSC4Rec., Comment: Published on KDD'25: Proceedings of the ACM SIGKDD Conference on Knowledge Discovery and Data Mining 2025

With the growing demand for Earth observation, it is important to provide reliable real-time remote sensing inference services to meet the low-latency requirements. The Space Computing Power Network (Space-CPN) offers a promising solution by providing onboard computing and extensive coverage capabilities for real-time inference. This paper presents a remote sensing artificial intelligence applications deployment framework designed for Low Earth Orbit satellite constellations to achieve real-time inference performance. The framework employs the microservice architecture, decomposing monolithic inference tasks into reusable, independent modules to address high latency and resource heterogeneity. This distributed approach enables optimized microservice deployment, minimizing resource utilization while meeting quality of service and functional requirements. We introduce Robust Optimization to the deployment problem to address data uncertainty. Additionally, we model the Robust Optimization problem as a Partially Observable Markov Decision Process and propose a robust reinforcement learning algorithm to handle the semi-infinite Quality of Service constraints. Our approach yields sub-optimal solutions that minimize accuracy loss while maintaining acceptable computational costs. Simulation results demonstrate the effectiveness of our framework., Comment: 14 pages

Machine learning techniques have opened new avenues for real-time quantum state tomography (QST). In this work, we demonstrate the deployment of machine learning-based QST onto edge devices, specifically utilizing field programmable gate arrays (FPGAs). This implementation is realized using the {\it Vitis AI Integrated Development Environment} provided by AMD\textsuperscript \textregistered~Inc. Compared to the Graphics Processing Unit (GPU)-based machine learning QST, our FPGA-based one reduces the average inference time by an order of magnitude, from 38 ms to 2.94 ms, but only sacrifices the average fidelity about $1\% $ reduction (from 0.99 to 0.98). The FPGA-based QST offers a highly efficient and precise tool for diagnosing quantum states, marking a significant advancement in the practical applications for quantum information processing and quantum sensing., Comment: 6 pages, 5 figures

The effective frequency of a mechanical resonator can be tuned via the spring effect induced by quadratic optomechanical (QOM) coupling, and both spontaneous symmetry breaking and anti-parity-time phase transition were predicted in the QOM systems. Here, we show that the mechanical susceptibility can be enhanced significantly by driving the QOM system with a strong external optical field, and divergence will happen as the driving strength approaches the critical point (CP) for spontaneous symmetry breaking. Based on the CP, we propose a highly sensitive temperature sensor with a mechanical resonator quadratically coupled to an optical mode. We find that the sensitivity of the temperature sensor can be enhanced by several orders of magnitude as the driving strength approaches the CP, and the sensitivity of the temperature sensor remains high in the low-temperature limit. Our work provides an effective way to realize highly sensitive temperature sensing at ultra-low temperature in the QOM systems., Comment: 9 pages, 4 figures

A long-term accurate and robust localization system is essential for mobile robots to operate efficiently outdoors. Recent studies have shown the significant advantages of the wheel-mounted inertial measurement unit (Wheel-IMU)-based dead reckoning system. However, it still drifts over extended periods because of the absence of external correction signals. To achieve the goal of long-term accurate localization, we propose Wheel-GINS, a Global Navigation Satellite System (GNSS)/inertial navigation system (INS) integrated navigation system using a Wheel-IMU. Wheel-GINS fuses the GNSS position measurement with the Wheel-IMU via an extended Kalman filter to limit the long-term error drift and provide continuous state estimation when the GNSS signal is blocked. Considering the specificities of the GNSS/Wheel-IMU integration, we conduct detailed modeling and online estimation of the Wheel-IMU installation parameters, including the Wheel-IMU leverarm and mounting angle and the wheel radius error. Experimental results have shown that Wheel-GINS outperforms the traditional GNSS/Odometer/INS integrated navigation system during GNSS outages. At the same time, Wheel-GINS can effectively estimate the Wheel-IMU installation parameters online and, consequently, improve the localization accuracy and practicality of the system. The source code of our implementation is publicly available (https://github.com/i2Nav-WHU/Wheel-GINS)., Comment: Accepted to IEEE Transactions on Intelligent Transportation Systems

In cloud-centric recommender system, regular data exchanges between user devices and cloud could potentially elevate bandwidth demands and privacy risks. On-device recommendation emerges as a viable solution by performing reranking locally to alleviate these concerns. Existing methods primarily focus on developing local adaptive parameters, while potentially neglecting the critical role of tailor-made model architecture. Insights from broader research domains suggest that varying data distributions might favor distinct architectures for better fitting. In addition, imposing a uniform model structure across heterogeneous devices may result in risking inefficacy on less capable devices or sub-optimal performance on those with sufficient capabilities. In response to these gaps, our paper introduces Forward-OFA, a novel approach for the dynamic construction of device-specific networks (both structure and parameters). Forward-OFA employs a structure controller to selectively determine whether each block needs to be assembled for a given device. However, during the training of the structure controller, these assembled heterogeneous structures are jointly optimized, where the co-adaption among blocks might encounter gradient conflicts. To mitigate this, Forward-OFA is designed to establish a structure-guided mapping of real-time behaviors to the parameters of assembled networks. Structure-related parameters and parallel components within the mapper prevent each part from receiving heterogeneous gradients from others, thus bypassing the gradient conflicts for coupled optimization. Besides, direct mapping enables Forward-OFA to achieve adaptation through only one forward pass, allowing for swift adaptation to changing interests and eliminating the requirement for on-device backpropagation. Experiments on real-world datasets demonstrate the effectiveness and efficiency of Forward-OFA., Comment: Accepted by KDD 2025

In this paper, we investigate the shadow and optical appearance of the hairy Kerr black hole illuminated by a thin accretion disk, the materials of which outside the innermost stable circular orbit (ISCO) move on the equatorial circular orbit, while inside the ISCO they quickly plunge into the black hole. The deformation parameter $\alpha$ and hair parameter $l_o$ are found to influence the motions of accretion as well as the redshift effect of the photon, such that they significantly affect the shadow and image of the hairy Kerr black hole. Especially, these two parameters have competing effects on the size of the black hole's shadow, and significantly increase the width of photon ring. This study provides a preliminary theoretical prediction that the image of the hairy Kerr black hole, especially the photon ring structure, may be used to constrain the hair parameters with future high-precision astronomical observation., Comment: 16 pages,7 figures

To address the growing size of AI model training data and the lack of a universal data selection methodology-factors that significantly drive up training costs -- this paper presents the General Information Metrics Evaluation (GIME) method. GIME leverages general information metrics from Objective Information Theory (OIT), including volume, delay, scope, granularity, variety, duration, sampling rate, aggregation, coverage, distortion, and mismatch to optimize dataset selection for training purposes. Comprehensive experiments conducted across diverse domains, such as CTR Prediction, Civil Case Prediction, and Weather Forecasting, demonstrate that GIME effectively preserves model performance while substantially reducing both training time and costs. Additionally, applying GIME within the Judicial AI Program led to a remarkable 39.56% reduction in total model training expenses, underscoring its potential to support efficient and sustainable AI development.

Scoring the Optical Character Recognition (OCR) capabilities of Large Multimodal Models (LMMs) has witnessed growing interest recently. Existing benchmarks have highlighted the impressive performance of LMMs in text recognition; however, their abilities on certain challenging tasks, such as text localization, handwritten content extraction, and logical reasoning, remain underexplored. To bridge this gap, we introduce OCRBench v2, a large-scale bilingual text-centric benchmark with currently the most comprehensive set of tasks (4x more tasks than the previous multi-scene benchmark OCRBench), the widest coverage of scenarios (31 diverse scenarios including street scene, receipt, formula, diagram, and so on), and thorough evaluation metrics, with a total of 10,000 human-verified question-answering pairs and a high proportion of difficult samples. After carefully benchmarking state-of-the-art LMMs on OCRBench v2, we find that 20 out of 22 LMMs score below 50 (100 in total) and suffer from five-type limitations, including less frequently encountered text recognition, fine-grained perception, layout perception, complex element parsing, and logical reasoning. The benchmark and evaluation scripts are available at https://github.com/Yuliang-liu/MultimodalOCR.

In this letter, we investigate the strong gravitational lensing effects around two classes of rotating regular black holes, which behave as non-singular Minkowski core at the center. Starting from the null geodesic in the equatorial plane of the regular black holes, we analyze the deflection angle for both prograde photons and retrograde photons, which are found to significantly shift from that for Kerr black hole. Then we suppose the rotating regular black holes as the supermassive M87* and SgrA* black holes, respectively, and evaluate the lensing observables such as the image position, separation, magnification and the time delays between the relativistic images for the black holes. In both cases, the time delay differences between the regular black holes and Kerr black hole seem to be measurable, especially for the M87* black hole, but the deviations of outermost relativistic image and the asymptotic relativistic image for the two rotating regular black holes from those for Kerr black hole are less than 10 $\mu as$. This means that it is difficult to probe the regularity of the black hole via the current observations, but we can expect the next generation Event Horizon Telescope to test more precise properties of strong field regime.

Observational constraints on small-scale primordial gravitational waves are considerably weaker than those on large scales. We focus on scenarios with significant primordial gravitational waves and curvature perturbations on small scales, studying the energy density spectrum of the second-order TSIGW. By leveraging current data from CMB, BAO, and PTA, combined with the SNR analysis of LISA, we can investigate how tensor-scalar induced gravitational waves affect observations on various scales, thus constraining the parameter space for primordial gravitational waves and curvature perturbations. The Bayes factor analysis suggests that TSIGW+PGW might be more likely to dominate current PTA observations compared to SMBHB., Comment: 19 pages, 7 figures

This paper investigates the observational appearances of an inner extremal regular black hole(IERBH) illuminated by various types of accretion models. The study reveals that when the BH is illuminated by specific accretion flows, the effects of quantum gravity become more pronounced,significantly impacting key observational features such as the shadow radius, photon ring, and total observed intensity. Specifically, the introduction of a more realistic radially infalling spherical accretion flow further accentuates these differences. This dynamic flow results in a darker central region in the BH image due to the Doppler effect, which modulates the observed intensity based on the relative motion of the infalling matter. The shadow radius and total observed intensity are notably affected by the quantum correction parameters, providing additional signatures that distinguish regular BHs from their classical counterparts.

Conformal prediction is a valuable tool for quantifying predictive uncertainty of machine learning models. However, its applicability relies on the assumption of data exchangeability, a condition which is often not met in real-world scenarios. In this paper, we consider the problem of adaptive conformal inference without any assumptions about the data generating process. Existing approaches for adaptive conformal inference are based on optimizing the pinball loss using variants of online gradient descent. A notable shortcoming of such approaches is in their explicit dependence on and sensitivity to the choice of the learning rates. In this paper, we propose a different approach for adaptive conformal inference that leverages parameter-free online convex optimization techniques. We prove that our method controls long-term miscoverage frequency at a nominal level and demonstrate its convincing empirical performance without any need of performing cumbersome parameter tuning.

We consider game-theoretically secure distributed protocols for coalition games that approximate the Shapley value with small multiplicative error. Since all known existing approximation algorithms for the Shapley value are randomized, it is a challenge to design efficient distributed protocols among mutually distrusted players when there is no central authority to generate unbiased randomness. The game-theoretic notion of maximin security has been proposed to offer guarantees to an honest player's reward even if all other players are susceptible to an adversary. Permutation sampling is often used in approximation algorithms for the Shapley value. A previous work in 1994 by Zlotkin et al. proposed a simple constant-round distributed permutation generation protocol based on commitment scheme, but it is vulnerable to rushing attacks. The protocol, however, can detect such attacks. In this work, we model the limited resources of an adversary by a violation budget that determines how many times it can perform such detectable attacks. Therefore, by repeating the number of permutation samples, an honest player's reward can be guaranteed to be close to its Shapley value. We explore both high probability and expected maximin security. We obtain an upper bound on the number of permutation samples for high probability maximin security, even with an unknown violation budget. Furthermore, we establish a matching lower bound for the weaker notion of expected maximin security in specific permutation generation protocols. We have also performed experiments on both synthetic and real data to empirically verify our results.

People who are blind or have low vision (BLV) encounter numerous challenges in their daily lives and work. To support them, various haptic assistive tools have been developed. Despite these advancements, the effective utilization of these tools -- including the optimal haptic feedback and on-body stimulation positions for different tasks along with their limitations -- remains poorly understood. Recognizing these gaps, we conducted a systematic literature review spanning two decades (2004-2024) to evaluate the development of haptic assistive tools within the HCI community. Our findings reveal that these tools are primarily used for understanding graphical information, providing guidance and navigation, and facilitating education and training, among other life and work tasks. We identified three main limitations: hardware limitations, functionality limitations, and UX and evaluation methods limitations. Based on these insights, we discuss potential research avenues and offer suggestions for enhancing the effectiveness of future haptic assistive technologies.

Log statements have become an integral part of modern software systems. Prior research efforts have focused on supporting the decisions of placing log statements, such as where/what to log, while automated generation or completion of log statements has received little attention. With the increasing use of Large Language Models (LLMs) for code-related tasks such as code completion or generation, automated methods for generating or completing log statements have gained much momentum. Fine-tuning open-source LLMs like the Llama series is often preferred by enterprises over using commercial ones like the GPT series due to considerations including privacy, security, openness, performance, etc. Fine-tuning LLMs requires task-specific training data and custom-designed processing algorithms, which, however, have not been thoroughly explored for the log statement generation task. This paper fills this gap by contributing such a fine-tuning method LoGFiLM and an exemplar model by using the proposed method to fine-tune Llama-3-8B. Experiments with our own curated dataset and a public dataset show that LoGFiLM consistently outperforms the original Llama-3-8B and the commercial LLMs of GPT-3.5 and GPT-4. The results further reveal that fine-tuning Llama-3-8B with data encompassing broader contextual ranges surrounding log statements yields a better model for the automated generation of log statements.

Federated learning (FL) is a promising technology for data privacy and distributed optimization, but it suffers from data imbalance and heterogeneity among clients. Existing FL methods try to solve the problems by aligning client with server model or by correcting client model with control variables. These methods excel on IID and general Non-IID data but perform mediocrely in Simpson's Paradox scenarios. Simpson's Paradox refers to the phenomenon that the trend observed on the global dataset disappears or reverses on a subset, which may lead to the fact that global model obtained through aggregation in FL does not accurately reflect the distribution of global data. Thus, we propose FedCFA, a novel FL framework employing counterfactual learning to generate counterfactual samples by replacing local data critical factors with global average data, aligning local data distributions with the global and mitigating Simpson's Paradox effects. In addition, to improve the quality of counterfactual samples, we introduce factor decorrelation (FDC) loss to reduce the correlation among features and thus improve the independence of extracted factors. We conduct extensive experiments on six datasets and verify that our method outperforms other FL methods in terms of efficiency and global model accuracy under limited communication rounds.

Tripartite entanglement as a remarkable resource in quantum information science has been extensively investigated in hybrid quantum systems, whereas it is generally weak and prone to be suppressed by noise, restricting its practical application in quantum technologies. Here, we propose how to enhance the tripartite entanglement among magnons, photons and phonons in a hybrid cavity-magnon optomechanical system by exploiting a synergistic effect of the optical parametric amplification (OPA) and mechanical parametric amplification (MPA). We find that in the case of individually applying the OPA or MPA, the tripartite entanglement can be enhanced by several folds. Remarkably, upon suitably tuning the phase matching of the two parametric fields presented simultaneously, the strengths of the entanglement can be further enhanced due to the constructive interference between the OPA and MPA. We also show that our method can improve the robustness of the entanglement against thermal noise. Our work provides a promising method for manipulating the entanglement with easy tunability and may serve as a useful tool for the enhancement and protection of fragile quantum resources., Comment: 12 pages, 4 figures, accepted for publication in PRA

We present ARCAS (Automated Root Cause Analysis System), a diagnostic platform based on a Domain Specific Language (DSL) built for fast diagnostic implementation and low learning curve. Arcas is composed of a constellation of automated troubleshooting guides (Auto-TSGs) that can execute in parallel to detect issues using product telemetry and apply mitigation in near-real-time. The DSL is tailored specifically to ensure that subject matter experts can deliver highly curated and relevant Auto-TSGs in a short time without having to understand how they will interact with the rest of the diagnostic platform, thus reducing time-to-mitigate and saving crucial engineering cycles when they matter most. This contrasts with platforms like Datadog and New Relic, which primarily focus on monitoring and require manual intervention for mitigation. ARCAS uses a Large Language Model (LLM) to prioritize Auto-TSGs outputs and take appropriate actions, thus suppressing the costly requirement of understanding the general behavior of the system. We explain the key concepts behind ARCAS and demonstrate how it has been successfully used for multiple products across Azure Synapse Analytics and Microsoft Fabric Synapse Data Warehouse., Comment: 13 pages, 6 figures

Dealing with missing data poses significant challenges in predictive analysis, often leading to biased conclusions when oversimplified assumptions about the missing data process are made. In cases where the data are missing not at random (MNAR), jointly modeling the data and missing data indicators is essential. Motivated by a real data application with partially missing multivariate outcomes related to leaf photosynthetic traits and several environmental covariates, we propose two methods under a selection model framework for handling data with missingness in the response variables suitable for recovering various missingness mechanisms. Both approaches use a multivariate extension of Bayesian additive regression trees (BART) to flexibly model the outcomes. The first approach simultaneously uses a probit regression model to jointly model the missingness. In scenarios where the relationship between the missingness and the data is more complex or non-linear, we propose a second approach using a probit BART model to characterize the missing data process, thereby employing two BART models simultaneously. Both models also effectively handle ignorable covariate missingness. The efficacy of both models compared to existing missing data approaches is demonstrated through extensive simulations, in both univariate and multivariate settings, and through the aforementioned application to the leaf photosynthetic trait data.

Ads demand forecasting for Walmart's ad products plays a critical role in enabling effective resource planning, allocation, and management of ads performance. In this paper, we introduce a comprehensive demand forecasting system that tackles hierarchical time series forecasting in business settings. Though traditional hierarchical reconciliation methods ensure forecasting coherence, they often trade off accuracy for coherence especially at lower levels and fail to capture the seasonality unique to each time-series in the hierarchy. Thus, we propose a novel framework "Multi-Stage Hierarchical Forecasting Reconciliation and Adjustment (Multi-Stage HiFoReAd)" to address the challenges of preserving seasonality, ensuring coherence, and improving accuracy. Our system first utilizes diverse models, ensembled through Bayesian Optimization (BO), achieving base forecasts. The generated base forecasts are then passed into the Multi-Stage HiFoReAd framework. The initial stage refines the hierarchy using Top-Down forecasts and "harmonic alignment." The second stage aligns the higher levels' forecasts using MinTrace algorithm, following which the last two levels undergo "harmonic alignment" and "stratified scaling", to eventually achieve accurate and coherent forecasts across the whole hierarchy. Our experiments on Walmart's internal Ads-demand dataset and 3 other public datasets, each with 4 hierarchical levels, demonstrate that the average Absolute Percentage Error from the cross-validation sets improve from 3% to 40% across levels against BO-ensemble of models (LGBM, MSTL+ETS, Prophet) as well as from 1.2% to 92.9% against State-Of-The-Art models. In addition, the forecasts at all hierarchical levels are proved to be coherent. The proposed framework has been deployed and leveraged by Walmart's ads, sales and operations teams to track future demands, make informed decisions and plan resources., Comment: Published in 2024 IEEE International Conference on Big Data (BigData)

Noisy labels are both inevitable and problematic in machine learning methods, as they negatively impact models' generalization ability by causing overfitting. In the context of learning with noise, the transition matrix plays a crucial role in the design of statistically consistent algorithms. However, the transition matrix is often considered unidentifiable. One strand of methods typically addresses this problem by assuming that the transition matrix is instance-independent; that is, the probability of mislabeling a particular instance is not influenced by its characteristics or attributes. This assumption is clearly invalid in complex real-world scenarios. To better understand the transition relationship and relax this assumption, we propose to study the data generation process of noisy labels from a causal perspective. We discover that an unobservable latent variable can affect either the instance itself, the label annotation procedure, or both, which complicates the identification of the transition matrix. To address various scenarios, we have unified these observations within a new causal graph. In this graph, the input instance is divided into a noise-resistant component and a noise-sensitive component based on whether they are affected by the latent variable. These two components contribute to identifying the ``causal transition matrix'', which approximates the true transition matrix with theoretical guarantee. In line with this, we have designed a novel training framework that explicitly models this causal relationship and, as a result, achieves a more accurate model for inferring the clean label.

Maximum structural chirality refers to the highest selectivity for circularly polarized light (CPL) in nanostructures, often manifested as maximum circular dichroism (CD), optical rotation (OR), and spin-orbit coupling (SOC). However, the underlying physical mechanisms that lead to maximum chirality remain unclear. In this work, we demonstrate that maximum chirality in dielectric nanostructures arises from the constructive and destructive interference of multipole moments with different CPL. By employing generalized multipole decomposition, we introduce a generalized chiral multipole mechanism that allows for direct numerical calculation of CD and establishes the conditions required to achieve maximum chirality. This approach provides a comprehensive framework for analyzing chirality and serves as a foundation for future investigations of chiral nanostructures., Comment: 11 pages, 6 figures

In this paper, a novel $h$-adaptive isogeometric solver utilizing high-order hierarchical splines is proposed to solve the all-electron Kohn--Sham equation. In virtue of the smooth nature of Kohn--Sham wavefunctions across the domain, except at the nuclear positions, high-order globally regular basis functions such as B-splines are well suited for achieving high accuracy. To further handle the singularities in the external potential at the nuclear positions, an $h$-adaptive framework based on the hierarchical splines is presented with a specially designed residual-type error indicator, allowing for different resolutions on the domain. The generalized eigenvalue problem raising from the discretized Kohn--Sham equation is effectively solved by the locally optimal block preconditioned conjugate gradient (LOBPCG) method with an elliptic preconditioner, and it is found that the eigensolver's convergence is independent of the spline basis order. A series of numerical experiments confirm the effectiveness of the $h$-adaptive framework, with a notable experiment that the numerical accuracy $10^{-3} \mathrm{~Hartree/particle}$ in the all-electron simulation of a methane molecule is achieved using only $6355$ degrees of freedom, demonstrating the competitiveness of our solver for the all-electron Kohn--Sham equation.

The separatrix operational space (SepOS) model [Eich \& Manz, \emph{Nuclear Fusion} (2021)] is shown to predict the L-H transition, the L-mode density limit, and the ideal MHD ballooning limit in terms of separatrix parameters for a wide range of Alcator C-Mod plasmas. The model is tested using Thomson scattering measurements across a wide range of operating conditions on C-Mod, spanning $\overline{n}_{e} = 0.3 - 5.5 \times 10^{20}$m$^{-3}$, $B_{t} = 2.5 - 8.0$ T, and $B_{p} = 0.1 - 1.2$ T. An empirical regression for the electron pressure gradient scale length, $\lambda_{p_{e}}$, against a turbulence control parameter, $\alpha_{t}$, and the poloidal fluid gyroradius, $\rho_{s,p}$, for H-modes is constructed and found to require positive exponents for both regression parameters, indicating turbulence widening of near-SOL widths at high $\alpha_{t}$ and an inverse scaling with $B_{p}$, consistent with results on AUG. The SepOS model is also tested in the unfavorable drift direction and found to apply well to all three boundaries, including the L-H transition as long as a correction to the Reynolds energy transfer term, $\alpha_\mathrm{RS} < 1$ is applied. I-modes typically exist in the unfavorable drift direction for values of $\alpha_{t} \lesssim 0.35$. Finally, an experiment studying the transition between the type-I ELMy and EDA H-mode is analyzed using the same framework. It is found that a recently identified boundary at $\alpha_{t} = 0.55$ excludes most EDA H-modes but that the balance of wavenumbers responsible for the L-mode density limit, namely $k_\mathrm{EM} = k_\mathrm{RBM}$, may better describe the transition on C-Mod. The ensemble of boundaries validated and explored is then applied to project regime access and limit avoidance for the SPARC primary reference discharge parameters.

As large language models (LLMs) process increasing context windows, the memory usage of KV cache has become a critical bottleneck during inference. The mainstream KV compression methods, including KV pruning and KV quantization, primarily focus on either token or precision dimension and seldom explore the efficiency of their combination. In this paper, we comprehensively investigate the token-precision trade-off in KV cache compression. Experiments demonstrate that storing more tokens in the KV cache with lower precision, i.e., quantized pruning, can significantly enhance the long-context performance of LLMs. Furthermore, in-depth analysis regarding token-precision trade-off from a series of key aspects exhibit that, quantized pruning achieves substantial improvements in retrieval-related tasks and consistently performs well across varying input lengths. Moreover, quantized pruning demonstrates notable stability across different KV pruning methods, quantization strategies, and model scales. These findings provide valuable insights into the token-precision trade-off in KV cache compression. We plan to release our code in the near future., Comment: 13pages,7 figures

Tabular data are fundamental in common machine learning applications, ranging from finance to genomics and healthcare. This paper focuses on tabular regression tasks, a field where deep learning (DL) methods are not consistently superior to machine learning (ML) models due to the challenges posed by irregular target functions inherent in tabular data, causing sensitive label changes with minor variations from features. To address these issues, we propose a novel Arithmetic-Aware Pre-training and Adaptive-Regularized Fine-tuning framework (APAR), which enables the model to fit irregular target function in tabular data while reducing the negative impact of overfitting. In the pre-training phase, APAR introduces an arithmetic-aware pretext objective to capture intricate sample-wise relationships from the perspective of continuous labels. In the fine-tuning phase, a consistency-based adaptive regularization technique is proposed to self-learn appropriate data augmentation. Extensive experiments across 10 datasets demonstrated that APAR outperforms existing GBDT-, supervised NN-, and pretrain-finetune NN-based methods in RMSE (+9.43% $\sim$ 20.37%), and empirically validated the effects of pre-training tasks, including the study of arithmetic operations. Our code and data are publicly available at https://github.com/johnnyhwu/APAR., Comment: AAAI 2025 Main Track

Domain Large Language Models (LLMs) are developed for domain-specific tasks based on general LLMs. But it still requires professional knowledge to facilitate the expertise for some domain-specific tasks. In this paper, we investigate into knowledge-intensive calculation problems. We find that the math problems to be challenging for LLMs, when involving complex domain-specific rules and knowledge documents, rather than simple formulations of terminologies. Therefore, we propose a pipeline to solve the domain-specific calculation problems with Knowledge-Intensive Programs Generator more effectively, named as KIPG. It generates knowledge-intensive programs according to the domain-specific documents. For each query, key variables are extracted, then outcomes which are dependent on domain knowledge are calculated with the programs. By iterative preference alignment, the code generator learns to improve the logic consistency with the domain knowledge. Taking legal domain as an example, we have conducted experiments to prove the effectiveness of our pipeline, and extensive analysis on the modules. We also find that the code generator is also adaptable to other domains, without training on the new knowledge., Comment: Under review

We demonstrate electronic sensing of DNA nanostar (NS) condensate. Specifically, we use electrokinetic nanofluidics to observe and interpret how temperature-induced NS condensation affects nanochannel current. The increase in current upon filling a nanochannel with NS condensate indicates that its electrophoretic mobility is about half that of a single NS and its effective ionic strength is $\sim35$\% greater than that of 150mM NaCl in phosphate buffer. $\zeta$-potential measurements before and after exposure to NS show that condensate binds the silica walls of a nanochannel more strongly than individual NS do under identical conditions. This binding increases electroosmotic flow, possibly enough to completely balance, or even exceed, the electrophoretic velocity of NS condensate. Although the current through a flat nanochannel is erratic in the presence of NS condensate, tilting the nanochannel to accumulate NS condensate at one entrance (and away from the other) results in a robust electronic signature of the NS phase transition at temperatures $T_c$ = $f$([NaCl]) that agree with those obtained by other methods. Electrokinetic nanofluidic detection and measurement of NS condensate thus provides a foundation for novel biosensing technologies based on liquid-liquid phase separation.

Metasurfaces are reshaping traditional optical paradigms and are increasingly required in complex applications that demand substantial computational resources to numerically solve Maxwell's equations-particularly for large-scale systems, inhomogeneous media, and densely packed metadevices. Conventional forward design using electromagnetic solvers is based on specific approximations, which may not effectively address complex problems. In contrast, existing inverse design methods are a stepwise process that is often time-consuming and involves repetitive computations. Here, we present an inverse design approach utilizing a surrogate Neighborhood Attention Transformer, MetaE-former, to predict the performance of metasurfaces with ultrafast speed and high accuracy. This method achieves global solutions for hundreds of nanostructures simultaneously, providing up to a 250,000-fold speedup compared with solving for individual meta-atoms based on the FDTD method. As examples, we demonstrate a binarized high-numerical-aperture (about 1.31) metalens and several optimized structured-light meta-generators. Our method significantly improves the beam shaping adaptability with metasurfaces and paves the way for fast designing of large-scale metadevices for shaping extreme light fields with high accuracy., Comment: 13 pages, 4 figures, 2 tables

In this work we investigate an inverse problem of recovering point sources and their time-dependent strengths from {a posteriori} partial internal measurements in a subdiffusion model which involves a Caputo fractional derivative in time and a general second-order elliptic operator in space. We establish the well-posedness of the direct problem in the sense of transposition and improved local regularity. Using classical unique continuation of the subdiffusion model and improved local solution regularity, we prove the uniqueness of simultaneously recovering the locations of point sources, time-dependent strengths and initial condition for both one- and multi-dimensional cases. Moreover, in the one-dimensional case, the elliptic operator can have time-dependent coefficients. These results extend existing studies on point source identification for parabolic type problems. Additionally we present several numerical experiments to show the feasibility of numerical reconstruction., Comment: 20 pages, 7 figures