Search

TianQin is a future space-based gravitational wave observatory targeting the frequency window of $10^{-4}$ Hz $\sim 1$ Hz. A large variety of gravitational wave sources are expected in this frequency band, including the merger of massive black hole binaries, the inspiral of extreme/intermediate mass ratio systems, stellar-mass black hole binaries, Galactic compact binaries, and so on. TianQin will consist of three Earth orbiting satellites on nearly identical orbits with orbital radii of about $10^5$ km. The satellites will form a normal triangle constellation whose plane is nearly perpendicular to the ecliptic plane. The TianQin project has been progressing smoothly following the ``0123" technology roadmap. In step ``0", the TianQin laser ranging station has been constructed and it has successfully ranged to all the five retro-reflectors on the Moon. In step ``1", the drag-free control technology has been tested and demonstrated using the TianQin-1 satellite. In step ``2", the inter-satellite laser interferometry technology will be tested using the pair of TianQin-2 satellites. The TianQin-2 mission has been officially approved and the satellites will be launched around 2026. In step ``3", i.e., the TianQin-3 mission, three identical satellites will be launched around 2035 to form the space-based gravitational wave detector, TianQin, and to start gravitational wave detection in space., Comment: 45 pages, 3 figures

An ideal multimodal agent should be aware of the quality of its input modalities. Recent advances have enabled large language models (LLMs) to incorporate auditory systems for handling various speech-related tasks. However, most audio LLMs remain unaware of the quality of the speech they process. This limitation arises because speech quality evaluation is typically excluded from multi-task training due to the lack of suitable datasets. To address this, we introduce the first natural language-based speech evaluation corpus, generated from authentic human ratings. In addition to the overall Mean Opinion Score (MOS), this corpus offers detailed analysis across multiple dimensions and identifies causes of quality degradation. It also enables descriptive comparisons between two speech samples (A/B tests) with human-like judgment. Leveraging this corpus, we propose an alignment approach with LLM distillation (ALLD) to guide the audio LLM in extracting relevant information from raw speech and generating meaningful responses. Experimental results demonstrate that ALLD outperforms the previous state-of-the-art regression model in MOS prediction, with a mean square error of 0.17 and an A/B test accuracy of 98.6%. Additionally, the generated responses achieve BLEU scores of 25.8 and 30.2 on two tasks, surpassing the capabilities of task-specific models. This work advances the comprehensive perception of speech signals by audio LLMs, contributing to the development of real-world auditory and sensory intelligent agents., Comment: ICLR 2025

Climate physical risks pose an increasing threat to urban infrastructure, necessitating urgent climate adaptation measures to protect lives and assets. Implementing such measures, including the development of resilient infrastructure and retrofitting existing systems, demands substantial financial investment. Unfortunately, due to the unprofitability stemming from the long-term returns, uncertainty, and complexity of infrastructure adaptation projects and the short-term profit-seeking objectives of private capital, a massive financial gap remains. This study suggests incentivizing private capital to bridge financial gaps through integrated carbon markets. Specifically, the framework combines carbon taxes and carbon markets to involve infrastructure and individuals in the climate mitigation phase, using the funds collected for climate adaptation. Moreover, it integrates lifestyle reformation, environmental mitigation, and infrastructure adaptation to establish harmonized standards and provide circular positive feedback to sustain the markets. We further explore how integrated carbon markets can facilitate fund collection and discuss the challenges of incorporating them into infrastructure climate adaptation. This study aims to foster collaboration between private and public capital to enable a more scientific, rational, and actionable implementation of integrated carbon markets, thus supporting sustainable financial backing for infrastructure climate adaptation, Comment: 18 pages,2 figures,99 references

Recently, end-to-end automatic speech recognition has become the mainstream approach in both industry and academia. To optimize system performance in specific scenarios, the Weighted Finite-State Transducer (WFST) is extensively used to integrate acoustic and language models, leveraging its capacity to implicitly fuse language models within static graphs, thereby ensuring robust recognition while also facilitating rapid error correction. However, WFST necessitates a frame-by-frame search of CTC posterior probabilities through autoregression, which significantly hampers inference speed. In this work, we thoroughly investigate the spike property of CTC outputs and further propose the conjecture that adjacent frames to non-blank spikes carry semantic information beneficial to the model. Building on this, we propose the Spike Window Decoding algorithm, which greatly improves the inference speed by making the number of frames decoded in WFST linearly related to the number of spiking frames in the CTC output, while guaranteeing the recognition performance. Our method achieves SOTA recognition accuracy with significantly accelerates decoding speed, proven across both AISHELL-1 and large-scale In-House datasets, establishing a pioneering approach for integrating CTC output with WFST., Comment: Accepted by ICASSP 2025

By analyzing $4.5$ fb$^{-1}$ of $e^{+}e^{-}$ collision data accumulated with the BESIII detector at center-of-mass energies ranging from $4599.53$ MeV to $4698.82$ MeV, we report the measurement of the absolute branching fraction (BF) of the inclusive decay $\Lambda_{c}^{+} \to K_{S}^{0} X$ using the double-tag technique. The result is $\mathcal{B}(\Lambda_{c}^{+} \to K_{S}^{0} X)=(10.9\pm0.2\pm0.1)\%$, where the first uncertainty is statistical and the second is systematic. This result indicates that there are still undiscovered decay channels containing $K_{S}^{0}$ in the final state with a combined BF of $(3.1\pm0.4)\%$. The BF of the inclusive decay $\Lambda_{c}^{+} \to \overline{K}^{0} / K^{0} X$ is calculated to be $\mathcal{B}(\Lambda_{c}^{+} \to \overline{K}^{0} / K^{0} X)=(21.8 \pm0.4 \pm0.2 \pm1.1)\%$, where the third uncertainty accounts for a possible difference between $\mathcal{B}(\Lambda_{c}^{+} \to K_{S}^{0} X)$ and $\mathcal{B}(\Lambda_{c}^{+} \to K_{L}^{0} X)$. The result is in agreement with the prediction of the statistical isospin model.

The emergence of foundation models in computational pathology has transformed histopathological image analysis, with whole slide imaging (WSI) diagnosis being a core application. Traditionally, weakly supervised fine-tuning via multiple instance learning (MIL) has been the primary method for adapting foundation models to WSIs. However, in this work we present a key experimental finding: a simple nonlinear mapping strategy combining mean pooling and a multilayer perceptron, called SiMLP, can effectively adapt patch-level foundation models to slide-level tasks without complex MIL-based learning. Through extensive experiments across diverse downstream tasks, we demonstrate the superior performance of SiMLP with state-of-the-art methods. For instance, on a large-scale pan-cancer classification task, SiMLP surpasses popular MIL-based methods by 3.52%. Furthermore, SiMLP shows strong learning ability in few-shot classification and remaining highly competitive with slide-level foundation models pretrained on tens of thousands of slides. Finally, SiMLP exhibits remarkable robustness and transferability in lung cancer subtyping. Overall, our findings challenge the conventional MIL-based fine-tuning paradigm, demonstrating that a task-agnostic representation strategy alone can effectively adapt foundation models to WSI analysis. These insights offer a unique and meaningful perspective for future research in digital pathology, paving the way for more efficient and broadly applicable methodologies., Comment: 11 pages, 3 figures, 4 tables

Machine Translation (MT) has been predominantly designed for sentence-level translation using transformer-based architectures. While next-token prediction based Large Language Models (LLMs) demonstrate strong capabilities in long-text translation, non-extensive language models often suffer from omissions and semantic inconsistencies when processing paragraphs. Existing preference alignment methods improve sentence-level translation but fail to ensure coherence over extended contexts due to the myopic nature of next-token generation. We introduce Plan2Align, a test-time alignment framework that treats translation as a predictive planning problem, adapting Model Predictive Control to iteratively refine translation outputs. Experiments on WMT24 Discourse-Level Literary Translation show that Plan2Align significantly improves paragraph-level translation, achieving performance surpassing or on par with the existing training-time and test-time alignment methods on LLaMA-3.1 8B., Comment: Preprint. Code will be released at Plan2Align GitHub link: https://github.com/NYCU-RL-Bandits-Lab/Plan2Align

Knowledge distillation (KD) aims to transfer the knowledge of a more capable yet cumbersome teacher model to a lightweight student model. In recent years, relation-based KD methods have fallen behind, as their instance-matching counterparts dominate in performance. In this paper, we revive relational KD by identifying and tackling several key issues in relation-based methods, including their susceptibility to overfitting and spurious responses. Specifically, we transfer novelly constructed affinity graphs that compactly encapsulate a wealth of beneficial inter-sample, inter-class, and inter-view correlations by exploiting virtual views and relations as a new kind of knowledge. As a result, the student has access to richer guidance signals and stronger regularisation throughout the distillation process. To further mitigate the adverse impact of spurious responses, we prune the affinity graphs by dynamically detaching redundant and unreliable edges. Extensive experiments on CIFAR-100 and ImageNet datasets demonstrate the superior performance of the proposed virtual relation matching (VRM) method over a range of models, architectures, and set-ups. For instance, VRM for the first time hits 74.0% accuracy for ResNet50-to-MobileNetV2 distillation on ImageNet, and improves DeiT-T by 14.44% on CIFAR-100 with a ResNet56 teacher. Thorough analyses are also conducted to gauge the soundness, properties, and complexity of our designs. Code and models will be released.

Understanding how neural networks process complex patterns of information is crucial for advancing both neuroscience and artificial intelligence. To investigate fundamental principles of neural computation, we studied dissociated neuronal cultures, one of the most primitive living neural networks, on high-resolution CMOS microelectrode arrays and tested whether the dissociated culture exhibits regularity sensitivity beyond mere stimulus-specific adaptation and deviance detection. In oddball electrical stimulation paradigms, we confirmed that the neuronal culture produced mismatch responses (MMRs) with true deviance detection beyond mere adaptation. These MMRs were dependent on the N-methyl-D-aspartate (NMDA) receptors, similar to mismatch negativity (MMN) in humans, which is known to have true deviance detection properties. Crucially, we also showed sensitivity to the statistical regularity of stimuli, a phenomenon previously observed only in intact brains: the MMRs in a predictable, periodic sequence were smaller than those in a commonly used sequence in which the appearance of the deviant stimulus was random and unpredictable. These results challenge the traditional view that a hierarchically structured neural network is required to process complex temporal patterns, suggesting instead that deviant detection and regularity sensitivity are inherent properties arising from the primitive neural network. They also suggest new directions for the development of neuro-inspired artificial intelligence systems, emphasizing the importance of incorporating adaptive mechanisms and temporal dynamics in the design of neural networks.

The {\Pi}-operator plays an important role in complex analysis, especially in the theory of generalized analytic functions in the sense of Vekua. In this paper, we introduce a generalized {\Pi}-operator in the theory of slice monogenic functions, and some mapping properties of the generalized {\Pi}-operator are also introduced. Further, a left and right inverse and the adjoint operator of the generalized {\Pi}-operator are given. As an application, we introduce a slice Beltrami equation, which reduces to the classical complex Beltrami equation when the dimension is 2. We show details that the norm estimate of the generalized {\Pi}-operator can determine the existence of solutions of the slice Beltrami equation.

The theory of generalized partial-slice monogenic functions is considered as a syhthesis of the classical Clifford analysis and the theory of slice monogenic functions. In this paper, we investigate the Cauchy integral formula and the Plemelj formula for generalized partial-slice monogenic functions. Further, we study some properties of the Teodorescu transform in this context. A norm estimation for the Teodorescu transform is discussed as well.

Advancements in Large Vision-Language Models (LVLMs) have demonstrated promising performance in a variety of vision-language tasks involving image-conditioned free-form text generation. However, growing concerns about hallucinations in LVLMs, where the generated text is inconsistent with the visual context, are becoming a major impediment to deploying these models in applications that demand guaranteed reliability. In this paper, we introduce a framework to address this challenge, ConfLVLM, which is grounded on conformal prediction to achieve finite-sample distribution-free statistical guarantees on the factuality of LVLM output. This framework treats an LVLM as a hypothesis generator, where each generated text detail (or claim) is considered an individual hypothesis. It then applies a statistical hypothesis testing procedure to verify each claim using efficient heuristic uncertainty measures to filter out unreliable claims before returning any responses to users. We conduct extensive experiments covering three representative application domains, including general scene understanding, medical radiology report generation, and document understanding. Remarkably, ConfLVLM reduces the error rate of claims generated by LLaVa-1.5 for scene descriptions from 87.8\% to 10.0\% by filtering out erroneous claims with a 95.3\% true positive rate. Our results further demonstrate that ConfLVLM is highly flexible, and can be applied to any black-box LVLMs paired with any uncertainty measure for any image-conditioned free-form text generation task while providing a rigorous guarantee on controlling the risk of hallucination.

High-quality benchmarks are essential for evaluating reasoning and retrieval capabilities of large language models (LLMs). However, curating datasets for this purpose is not a permanent solution as they are prone to data leakage and inflated performance results. To address these challenges, we propose PhantomWiki: a pipeline to generate unique, factually consistent document corpora with diverse question-answer pairs. Unlike prior work, PhantomWiki is neither a fixed dataset, nor is it based on any existing data. Instead, a new PhantomWiki instance is generated on demand for each evaluation. We vary the question difficulty and corpus size to disentangle reasoning and retrieval capabilities respectively, and find that PhantomWiki datasets are surprisingly challenging for frontier LLMs. Thus, we contribute a scalable and data leakage-resistant framework for disentangled evaluation of reasoning, retrieval, and tool-use abilities. Our code is available at https://github.com/kilian-group/phantom-wiki.

We identify observational signatures suggesting a history of dynamical instability in 26 out of 34 M-dwarf multi-planet systems containing no large planets. These systems may have primarily formed in a gas-rich environment, potentially hosted more planets and were more compact. We extend previous simulations of the formation of the TRAPPIST-1 system to 100 Myr to test the stability of these systems without gas. We find the absence of a strong mean motion resonance in the innermost planet pair and the absence of three body resonances throughout the system are likely to result in the merging and ejection of planets after the gas disk disperses. The runs that experience such an instability tend to produce final systems with lower multiplicities, period ratios larger than two, increased orbital spacings, higher planetary angular momentum deficits, and slightly smaller mass ratios between adjacent planets. Remarkably, we find these same trends in the observations of M-dwarf multi-planet systems containing no large planets. Our work allows us to identify specific systems that may have experienced an instability and suggests that only ~25% of these systems formed in their current observed state while most systems were likely more compact and multiplicitous earlier in time. Previous research indicates that systems that have experienced a late stage giant impact may host planets potentially more habitable than the systems that did not. With this in mind, we suggest systems around M-dwarfs that contain period ratios larger than two be given priority in the search for habitable worlds., Comment: 16 pages, 5 figures, and 3 tables. Accepted for publication in ApJ

Decentralized Federated Learning (DFL) is an emerging paradigm that enables collaborative model training without centralized data aggregation, enhancing privacy and resilience. However, its sustainability remains underexplored, as energy consumption and carbon emissions vary across different system configurations. Understanding the environmental impact of DFL is crucial for optimizing its design and deployment. This study aims to assess the sustainability of DFL systems by analyzing factors that influence energy consumption and carbon emissions. Additionally, it proposes sustainability-aware optimization strategies, including a node selection algorithm and an aggregation method, to reduce the environmental footprint of DFL without compromising model performance. The proposed framework, named GreenDFL, systematically evaluates the impact of hardware accelerators, model architecture, communication medium, data distribution, network topology, and federation size on sustainability. Empirical experiments are conducted on multiple datasets using different system configurations, measuring energy consumption and carbon emissions across various phases of the DFL lifecycle. Results indicate that local training dominates energy consumption and carbon emissions, while communication has a relatively minor impact. Optimizing model complexity, using GPUs instead of CPUs, and strategically selecting participating nodes significantly improve sustainability. Additionally, deploying nodes in regions with lower carbon intensity and integrating early stopping mechanisms further reduce emissions. The proposed framework provides a comprehensive and practical computational approach for assessing the sustainability of DFL systems. Furthermore, it offers best practices for improving environmental efficiency in DFL, making sustainability considerations more actionable in real-world deployments.

Mainstream issue-resolving frameworks predominantly rely on commercial models, leading to high costs and privacy concerns. Existing training approaches for issue resolving struggle with poor generalization and fail to fully leverage open-source development resources. We propose Subtask-oriented Reinforced Fine-Tuning (SoRFT), a novel training approach to enhance the issue resolving capability of LLMs. We decomposes issue resolving into structured subtasks: file localization, function localization, line localization, and code edit generation. SoRFT consists of two training stages: (1) rejection-sampled supervised fine-tuning, Chain of Thought (CoT) data is filtered using ground-truth before fine-tuning the LLM, and (2) rule-based reinforcement learning, which leverages PPO with ground-truth based rewards. We evaluate the SoRFT-trained model on SWE-Bench Verified and SWE-Bench Lite, achieving state-of-the-art (SOTA) performance among open-source models (e.g., resolve 21.4% issues on SWE-Bench Verified with SoRFT-Qwen-7B). The experimental results demonstrate that SoRFT significantly enhances issue-resolving performance, improves model generalization, and provides a cost-efficient alternative to commercial models.

In the post-Moore era, the need for efficient solutions to non-deterministic polynomial-time (NP) problems is becoming more pressing. In this context, the Ising model implemented by the probabilistic computing systems with probabilistic bits (p-bits) has attracted attention due to the widespread availability of p-bits and support for large-scale simulations. This study marks the first work to apply probabilistic computing to tackle protein folding, a significant NP-complete problem challenge in biology. We represent proteins as sequences of hydrophobic (H) and polar (P) beads within a three-dimensional (3-D) grid and introduce a novel many-body interaction-based encoding method to map the problem onto an Ising model. Our simulations show that this approach significantly simplifies the energy landscape for short peptide sequences of six amino acids, halving the number of energy levels. Furthermore, the proposed mapping method achieves approximately 100 times acceleration for sequences consisting of ten amino acids in identifying the correct folding configuration. We predicted the optimal folding configuration for a peptide sequence of 36 amino acids by identifying the ground state. These findings highlight the unique potential of the proposed encoding method for solving protein folding and, importantly, provide new tools for solving similar NP-complete problems in biology by probabilistic computing approach., Comment: 14pages, 6 fingures

Combining multiple perceptual inputs and performing combinatorial reasoning in complex scenarios is a sophisticated cognitive function in humans. With advancements in multi-modal large language models, recent benchmarks tend to evaluate visual understanding across multiple images. However, they often overlook the necessity of combinatorial reasoning across multiple perceptual information. To explore the ability of advanced models to integrate multiple perceptual inputs for combinatorial reasoning in complex scenarios, we introduce two benchmarks: Clue-Visual Question Answering (CVQA), with three task types to assess visual comprehension and synthesis, and Clue of Password-Visual Question Answering (CPVQA), with two task types focused on accurate interpretation and application of visual data. For our benchmarks, we present three plug-and-play approaches: utilizing model input for reasoning, enhancing reasoning through minimum margin decoding with randomness generation, and retrieving semantically relevant visual information for effective data integration. The combined results reveal current models' poor performance on combinatorial reasoning benchmarks, even the state-of-the-art (SOTA) closed-source model achieves only 33.04% accuracy on CVQA, and drops to 7.38% on CPVQA. Notably, our approach improves the performance of models on combinatorial reasoning, with a 22.17% boost on CVQA and 9.40% on CPVQA over the SOTA closed-source model, demonstrating its effectiveness in enhancing combinatorial reasoning with multiple perceptual inputs in complex scenarios. The code will be publicly available., Comment: 11pages

Quantum error correction is crucial for large-scale quantum computing, but the absence of efficient decoders for new codes like quantum low-density parity-check (QLDPC) codes has hindered progress. Here we introduce a universal decoder based on linear attention sequence modeling and graph neural network that operates directly on any stabilizer code's graph structure. Our numerical experiments demonstrate that this decoder outperforms specialized algorithms in both accuracy and speed across diverse stabilizer codes, including surface codes, color codes, and QLDPC codes. The decoder maintains linear time scaling with syndrome measurements and requires no structural modifications between different codes. For the Bivariate Bicycle code with distance 12, our approach achieves a 39.4% lower logical error rate than previous best decoders while requiring only ~1% of the decoding time. These results provide a practical, universal solution for quantum error correction, eliminating the need for code-specific decoders.

We establish a new type of weak Harnack estimates with optimal parabolic tail for the weak supersolutions to a doubly nonlinear nonlocal $p$-Laplace equation, which is modeled on the nonlocal Trudinger equation. Our results are achieved by employing the expansion of positivity and measure theoretical techniques. In particular, the weak Harnack estimates highlight the nonlocal feature, as we only require the local positivity of weak supersolutions instead of the global one.

Using $(2259.3 \pm 11.1)\times10^{6}$ $\psi(2S)$ events acquired with the BESIII detector, the branching fraction of $\psi(2S)\rightarrow\tau^{+}\tau^{-}$ is measured with improved precision to be $\mathcal{B}_{\psi(2S)\rightarrow\tau^{+}\tau^{-}}=(3.240~\pm~0.023~\pm~0.081)\times 10^{-3}$, where the first and second uncertainties are statistical and systematic, respectively, which is consistent with the world average value within one standard deviation. This value, along with those for the branching fractions of the $\psi(2S)$ decaying into $e^{+}e^{-}$ and $\mu^{+}\mu^{-}$, is in good agreement with the relation predicted by the sequential lepton hypothesis. Combining the branching fraction values with the leptonic width of the $\psi(2S)$, the total width of the $\psi(2S)$ is determined to be (287 $\pm$ 9) keV., Comment: 10 page, 5 figures

Previous successful approaches to missing modality completion rely on carefully designed fusion techniques and extensive pre-training on complete data, which can limit their generalizability in out-of-domain (OOD) scenarios. In this study, we pose a new challenge: can we develop a missing modality completion model that is both resource-efficient and robust to OOD generalization? To address this, we present a training-free framework for missing modality completion that leverages large multimodal models (LMMs). Our approach, termed the "Knowledge Bridger", is modality-agnostic and integrates generation and ranking of missing modalities. By defining domain-specific priors, our method automatically extracts structured information from available modalities to construct knowledge graphs. These extracted graphs connect the missing modality generation and ranking modules through the LMM, resulting in high-quality imputations of missing modalities. Experimental results across both general and medical domains show that our approach consistently outperforms competing methods, including in OOD generalization. Additionally, our knowledge-driven generation and ranking techniques demonstrate superiority over variants that directly employ LMMs for generation and ranking, offering insights that may be valuable for applications in other domains., Comment: Accepted to CVPR 2025

The tractor-trailer vehicle (robot) consists of a drivable tractor and one or more non-drivable trailers connected via hitches. Compared to typical car-like robots, the addition of trailers provides greater transportation capability. However, this also complicates motion planning due to the robot's complex kinematics, high-dimensional state space, and deformable structure. To efficiently plan safe, time-optimal trajectories that adhere to the kinematic constraints of the robot and address the challenges posed by its unique features, this paper introduces a lightweight, compact, and high-order smooth trajectory representation for tractor-trailer robots. Based on it, we design an efficiently solvable spatio-temporal trajectory optimization problem. To deal with deformable structures, which leads to difficulties in collision avoidance, we fully leverage the collision-free regions of the environment, directly applying deformations to trajectories in continuous space. This approach not requires constructing safe regions from the environment using convex approximations through collision-free seed points before each optimization, avoiding the loss of the solution space, thus reducing the dependency of the optimization on initial values. Moreover, a multi-terminal fast path search algorithm is proposed to generate the initial values for optimization. Extensive simulation experiments demonstrate that our approach achieves several-fold improvements in efficiency compared to existing algorithms, while also ensuring lower curvature and trajectory duration. Real-world experiments involving the transportation, loading and unloading of goods in both indoor and outdoor scenarios further validate the effectiveness of our method. The source code is accessible at https://github.com/ZJU-FAST-Lab/tracailer/., Comment: 15 pages, 12 figures

Time-independent methods, such as the transfer matrix method, are widely used to analyze the scattering properties of non-Hermitian systems. However, we demonstrate that these methods become invalid when the scattering matrix (S-matrix) exhibits poles in the first quadrant of the complex wave-number plane, indicating the presence of time-growing bound states within the system. The breakdown of time-independent approaches is attributed to their inherent omission of these bound states. We illustrate this using tight-binding models where non-Hermiticity is introduced through imaginary on-site potentials or asymmetric hopping terms. In all the models considered, parameter regimes exist where time-independent methods fail. Our findings highlight the critical importance of examining the distribution of S-matrix poles when applying time-independent methods to non-Hermitian scattering systems. Inappropriate application of these methods can lead to unphysical results and erroneous conclusions., Comment: 12 pages, 5 figures

We compare two approaches for solving high-dimensional eigenvalue problems with low-rank structure: the inexact Lanczos method and inexact polynomial-filtered subspace iteration. Inexactness stems from low-rank compression, enabling efficient representation of high-dimensional vectors in a low-rank tensor format. A primary challenge in these methods is that standard operations, such as matrix-vector products and linear combinations, increase tensor rank, necessitating rank truncation and hence approximation. The Lanczos method constructs an approximate orthonormal Krylov basis, which is often difficult to represent accurately using low-rank tensor formats, even when the eigenvectors themselves exhibit low-rank structure. In contrast, the low-rank polynomial-filtered subspace iteration uses approximate eigenvectors (Ritz vectors) directly as a subspace basis, bypassing the need for an orthonormal Krylov basis. Our analysis and numerical experiments demonstrate that inexact subspace iteration is much more robust to rank-truncation errors compared to the inexact Lanczos method. We further demonstrate that rank-truncated subspace iteration can converge for problems where the density matrix renormalization group method (DMRG) stagnates., Comment: 26 pages, 7 figures

This paper introduces a novel approach to demand estimation that utilizes partial observations of segment-level track counts. Building on established simulation-based demand estimation methods, we present a modified formulation that integrates sample track counts as a regularization term. This approach effectively addresses the underdetermination challenge in demand estimation, moving beyond the conventional reliance on a prior OD matrix. The proposed formulation aims to preserve the distribution of the observed track counts while optimizing the demand to align with observed path-level travel times. We tested this approach on Seattle's highway network with various congestion levels. Our findings reveal significant enhancements in the solution quality, particularly in accurately recovering ground truth demand patterns at both the OD and segment levels., Comment: Published in the 12th Triennial Symposium on Transportation Analysis conference (TRISTAN XII)

Label-free tomographic microscopy offers a compelling means to visualize three-dimensional (3D) refractive index (RI) distributions from two-dimensional (2D) intensity measurements. However, limited forward-model accuracy and the ill-posed nature of the inverse problem hamper artifact-free reconstructions. Meanwhile, artificial neural networks excel at modeling nonlinearities. Here, we employ a Differentiable Imaging framework that represents the 3D sample as a multi-layer neural network embedding physical constraints of light propagation. Building on this formulation, we propose a physics-guided Adaptive Dropout Neural Network (ADNN) for optical diffraction tomography (ODT), focusing on network topology and voxel-wise RI fidelity rather than solely on input-output mappings. By exploiting prior knowledge of the sample's RI, the ADNN adaptively drops and reactivates neurons, enhancing reconstruction accuracy and stability. We validate this method with extensive simulations and experiments on weakly and multiple-scattering samples under different imaging setups. The ADNN significantly improves quantitative 3D RI reconstructions, providing superior optical-sectioning and effectively suppressing artifacts. Experimental results show that the ADNN reduces the Mean Absolute Error (MAE) by a factor of 3 to 5 and increases the Structural Similarity Index Metric (SSIM) by about 4 to 30 times compared to the state-of-the-art approach., Comment: 30 pages, 17 figures

Incomplete multi-view clustering has become one of the important research problems due to the extensive missing multi-view data in the real world. Although the existing methods have made great progress, there are still some problems: 1) most methods cannot effectively mine the information hidden in the missing data; 2) most methods typically divide representation learning and clustering into two separate stages, but this may affect the clustering performance as the clustering results directly depend on the learned representation. To address these problems, we propose a novel incomplete multi-view clustering method with hierarchical information transfer. Firstly, we design the view-specific Graph Convolutional Networks (GCN) to obtain the representation encoding the graph structure, which is then fused into the consensus representation. Secondly, considering that one layer of GCN transfers one-order neighbor node information, the global graph propagation with the consensus representation is proposed to handle the missing data and learn deep representation. Finally, we design a weight-sharing pseudo-classifier with contrastive learning to obtain an end-to-end framework that combines view-specific representation learning, global graph propagation with hierarchical information transfer, and contrastive clustering for joint optimization. Extensive experiments conducted on several commonly-used datasets demonstrate the effectiveness and superiority of our method in comparison with other state-of-the-art approaches. The code is available at https://github.com/KelvinXuu/GHICMC.

The upcoming generation of mobile telecommunication systems is expected to support new use cases, where the mobile network serves one or more IP subnetworks located behind the User Equipment (UEs). This would create new challenges for the mobile system to efficiently serve such behind-UE subnetworks, as the latter are commonly not visible to the mobile system. In 3GPP, there have been works on Time-Sensitive Networking (TSN) and Deterministic Networking (DetNet), where the 5G System (5GS) is considered as a bridge or a DetNet node. In order to efficiently serve behind-UE IP subnetworks, we foresee the need for a further generalization where the mobile system (5GS and beyond) acts as a set of IP routers with more generic capabilities. In this article, we introduce the concept of Mobile System Router (MS-Router) which aims to provide a reference architectural design to enable the support of IP routing in the next generation of mobile telecommunication systems. The concept models a mobile system as an IP router per User Plane granularity. Each MS-Router implements an IP routing protocol, exchanges routing messages with the external routers and constructs a routing table, enabling the mobile system to dynamically learn the topology of the IP subnetwork behind the UEs and Data Network. The learned topology is translated into User Plane configuration to serve the IP subnetworks in an optimal way. The article also advocates different approaches where routing protocols can be implemented in the mobile system.

Let $\theta(z),\varphi(w)$ be two nonconstant inner functions and $M$ be a submodule in $H^2(\mathbb{D}^2)$. Let $C_{\theta,\varphi}$ denote the composition operator on $H^2(\mathbb{D}^2)$ defined by $C_{\theta,\varphi}f(z,w)=f(\theta(z),\varphi(w))$, and $M_{\theta,\varphi}$ denote the submodule $[C_{\theta,\varphi}M]$, that is, the smallest submodule containing $C_{\theta,\varphi}M$. Let $K^M_{\lambda,\mu}(z,w)$ and $K^{M_{\theta,\varphi}}_{\lambda,\mu}(z,w)$ be the reproducing kernel of $M$ and $M_{\theta,\varphi}$, respectively. By making full use of the positivity of certain de Branges-Rovnyak kernels, we prove that \[K^{M_{\theta,\varphi}}= K^M \circ B~ \cdot R,\] where $B=(\theta,\varphi)$, $R_{\lambda,\mu}(z,w)=\frac{1-\overline{\theta(\lambda)}\theta(z)}{1-\bar{\lambda}z} \frac{1-\overline{\varphi(\mu)}\varphi(w)}{1-\bar{\mu}w}$. This implies that $M_{\theta,\varphi}$ is a Hilbert-Schmidt submodule if and only if $M$ is. Moreover, as an application, we prove that the Hilbert-Schmidt norms of submodules $[\theta(z)-\varphi(w)]$ are uniformly bounded.

In recent years, Large Language Models (LLMs) have faced increasing demands to selectively remove sensitive information, protect privacy, and comply with copyright regulations through unlearning, by Machine Unlearning. While evaluating unlearning effectiveness is crucial, existing benchmarks are limited in scale and comprehensiveness, typically containing only a few hundred test cases. We identify two critical challenges in generating holistic audit datasets: ensuring audit adequacy and handling knowledge redundancy between forget and retain dataset. To address these challenges, we propose HANKER, an automated framework for holistic audit dataset generation leveraging knowledge graphs to achieve fine-grained coverage and eliminate redundant knowledge. Applying HANKER to the popular MUSE benchmark, we successfully generated over 69,000 and 111,000 audit cases for the News and Books datasets respectively, identifying thousands of knowledge memorization instances that the previous benchmark failed to detect. Our empirical analysis uncovers how knowledge redundancy significantly skews unlearning effectiveness metrics, with redundant instances artificially inflating the observed memorization measurements ROUGE from 19.7% to 26.1% and Entailment Scores from 32.4% to 35.2%, highlighting the necessity of systematic deduplication for accurate assessment., Comment: 11 pages, 4 figures

Using radial velocity to estimate binary fraction and predict orbital parameter distribution laws is an important method in the studies of binary stars. However, the high dimensionality of the orbital parameters of the binary system pose challenges to this work, particularly in terms of sample size and the accuracy of the calculated results. We have proposed and implemented an algorithm called DVCD (Differential Velocity Cumulative Distribution) for the statistical analysis of binary properties. % On the basis of higher accuracy than previous methods, the running efficiency of the method is greatly improved. This method not only achieves higher accuracy than previous approaches but also significantly improves computational efficiency. The computing time for the same sample and the hardware environment was reduced to $10^{-4}$ to $10^{-5}$ of that of some previous methods. Red giant samples from APOGEE DR16 were analyzed using the DVCD method. Both the hyperparameter of the orbital period $\pi$ and the intrinsic binary fraction ($f_{bin}$) can be constrained. We divided the sample into 16 subsets based on log$g$ and $M/H$. Calculation results showed that the binary fraction decreases with the decrease of surface gravity and decreases with the increase of metallicity. This trend provides important constraints on the evolution of binary stars., Comment: 22pages, 14 figures

The softmax function is crucial in Transformer attention, which normalizes each row of the attention scores with summation to one, achieving superior performances over other alternative functions. However, the softmax function can face a gradient vanishing issue when some elements of the attention scores approach extreme values, such as probabilities close to one or zero. In this paper, we propose Self-Adjust Softmax (SA-Softmax) to address this issue by modifying $softmax(x)$ to $x \cdot softmax(x)$ and its normalized variant $\frac{(x - min(x_{\min},0))}{max(0,x_{max})-min(x_{min},0)} \cdot softmax(x)$. We theoretically show that SA-Softmax provides enhanced gradient properties compared to the vanilla softmax function. Moreover, SA-Softmax Attention can be seamlessly integrated into existing Transformer models to their attention mechanisms with minor adjustments. We conducted experiments to evaluate the empirical performance of Transformer models using SA-Softmax compared to the vanilla softmax function. These experiments, involving models with up to 2.7 billion parameters, are conducted across diverse datasets, language tasks, and positional encoding methods., Comment: Tech Report

The multiferroic materials, which coexist magnetism, ferroelectric, and ferrovalley, have broad practical application prospects in promoting the miniaturization and integration of spintronic and valleytronic devices. However, it is rare that there are triferroic orders and multiple types of valley polarization in a real material. Here, we propose a mechanism to realize triferroic order coexistence and multiple types of valley polarization by structural phase transition in two-dimensional (2D) materials. The 1T and 2H phase OsBr2 monolayers exhibit non-magnetic semiconductor and ferromagnetic semiconductor with valley polarization up to 175.49 meV, respectively. Interestingly, the 1T phase OsBr2 bilayer shows the tri-state valley polarization due to lattice symmetry breaking, while the valley polarization of 2H phase bilayer originates from the combined effect of time-reversal symmetry breaking and spin-orbit coupling. Furthermore, the valley polarization and ferroelectric polarization of 1T phase AB stackings and 2H phase AA stackings can be manipulated via interlayer sliding. Importantly, we have verified that the 2H phase can be transformed to 1T phase by Li+ ion intercalation, while the 2H phase can occur the structural phase transition into the 1T phase by infrared laser induction. Our work provides a feasible strategy for manipulating valley polarization and a design idea for nano-devices with nonvolatile multiferroic properties., Comment: 10 pages, 7 figures

Altermagnetism, as a new category of collinear magnetism distinct from traditional ferromagnetism and antiferromagnetism, exhibits the spin splitting without net magnetization. Currently, researchers are focus on searching three-dimensional altermagnetism and exploring its novel physical properties. However, there is a lack of understanding of the physical origin of two-dimensional altermagnetic emergent behavior. Here, we propose an approach to realize the transition from Neel antiferromagnetism to altermagnetism in two-dimensional system using an electric field, Janus structure, and ferroelectric substrate. In monolayer VPSe3, we demonstrate that multiple-physical-fields cause the upper and lower Se atoms unequal to break PT symmetry, resulting in altermagnetic spin splitting. Noted that monolayer VPSe3 produces a spontaneous valley splitting of 2.91 meV at the conduction band minimum. The electric field can effectively tune the valley splitting magnitude, while the Janus structure not only changes the valley splitting magnitude, but also alters the direction. More interestingly, when the ferroelectric polarization of Al2S3 is upward, the direction of valley polarization is switched and the magnitude is almost unchanged. However, the valley splitting sigfinicantly increases under the downward. It is worth noting that the ferroelectric polarization can switch altermagnetic effect and realize anomalous valley Hall effect. Besides, we reveal the microscopic mechanism of valley splitting by an effective Hamiltonian. Our findings not only provide a method to designing altermagnet, but also enriches the valley physics., Comment: 8 pages, 5 figures

Achieving large-scale aerial swarms is challenging due to the inherent contradictions in balancing computational efficiency and scalability. This paper introduces Primitive-Swarm, an ultra-lightweight and scalable planner designed specifically for large-scale autonomous aerial swarms. The proposed approach adopts a decentralized and asynchronous replanning strategy. Within it is a novel motion primitive library consisting of time-optimal and dynamically feasible trajectories. They are generated utlizing a novel time-optimial path parameterization algorithm based on reachability analysis (TOPP-RA). Then, a rapid collision checking mechanism is developed by associating the motion primitives with the discrete surrounding space according to conflicts. By considering both spatial and temporal conflicts, the mechanism handles robot-obstacle and robot-robot collisions simultaneously. Then, during a replanning process, each robot selects the safe and minimum cost trajectory from the library based on user-defined requirements. Both the time-optimal motion primitive library and the occupancy information are computed offline, turning a time-consuming optimization problem into a linear-complexity selection problem. This enables the planner to comprehensively explore the non-convex, discontinuous 3-D safe space filled with numerous obstacles and robots, effectively identifying the best hidden path. Benchmark comparisons demonstrate that our method achieves the shortest flight time and traveled distance with a computation time of less than 1 ms in dense environments. Super large-scale swarm simulations, involving up to 1000 robots, running in real-time, verify the scalability of our method. Real-world experiments validate the feasibility and robustness of our approach. The code will be released to foster community collaboration., Comment: Accepted by IEEE Transactions on Robotics

Semantic communication marks a new paradigm shift from bit-wise data transmission to semantic information delivery for the purpose of bandwidth reduction. To more effectively carry out specialized downstream tasks at the receiver end, it is crucial to define the most critical semantic message in the data based on the task or goal-oriented features. In this work, we propose a novel goal-oriented communication (GO-COM) framework, namely Goal-Oriented Semantic Variational Autoencoder (GOS-VAE), by focusing on the extraction of the semantics vital to the downstream tasks. Specifically, we adopt a Vector Quantized Variational Autoencoder (VQ-VAE) to compress media data at the transmitter side. Instead of targeting the pixel-wise image data reconstruction, we measure the quality-of-service at the receiver end based on a pre-defined task-incentivized model. Moreover, to capture the relevant semantic features in the data reconstruction, imitation learning is adopted to measure the data regeneration quality in terms of goal-oriented semantics. Our experimental results demonstrate the power of imitation learning in characterizing goal-oriented semantics and bandwidth efficiency of our proposed GOS-VAE., Comment: Accepted for publication in 2025 International Conference on Communications (IEEE ICC); 6 pages, 4 figures

We present LAM, an innovative Large Avatar Model for animatable Gaussian head reconstruction from a single image. Unlike previous methods that require extensive training on captured video sequences or rely on auxiliary neural networks for animation and rendering during inference, our approach generates Gaussian heads that are immediately animatable and renderable. Specifically, LAM creates an animatable Gaussian head in a single forward pass, enabling reenactment and rendering without additional networks or post-processing steps. This capability allows for seamless integration into existing rendering pipelines, ensuring real-time animation and rendering across a wide range of platforms, including mobile phones. The centerpiece of our framework is the canonical Gaussian attributes generator, which utilizes FLAME canonical points as queries. These points interact with multi-scale image features through a Transformer to accurately predict Gaussian attributes in the canonical space. The reconstructed canonical Gaussian avatar can then be animated utilizing standard linear blend skinning (LBS) with corrective blendshapes as the FLAME model did and rendered in real-time on various platforms. Our experimental results demonstrate that LAM outperforms state-of-the-art methods on existing benchmarks.

Speculative decoding has become a promising technique to mitigate the high inference latency of autoregressive decoding in Large Language Models (LLMs). Despite its promise, the effective application of speculative decoding in LLMs still confronts three key challenges: the increasing memory demands of the draft model, the distribution shift between the short-training corpora and long-context inference, and inefficiencies in attention implementation. In this work, we enhance the performance of speculative decoding in long-context settings by addressing these challenges. First, we propose a memory-efficient draft model with a constant-sized Key-Value (KV) cache. Second, we introduce novel position indices for short-training data, enabling seamless adaptation from short-context training to long-context inference. Finally, we present an innovative attention aggregation method that combines fast implementations for prefix computation with standard attention for tree mask handling, effectively resolving the latency and memory inefficiencies of tree decoding. Our approach achieves strong results on various long-context tasks, including repository-level code completion, long-context summarization, and o1-like long reasoning tasks, demonstrating significant improvements in latency reduction. The code is available at https://github.com/sail-sg/LongSpec.

We present X-Dancer, a novel zero-shot music-driven image animation pipeline that creates diverse and long-range lifelike human dance videos from a single static image. As its core, we introduce a unified transformer-diffusion framework, featuring an autoregressive transformer model that synthesize extended and music-synchronized token sequences for 2D body, head and hands poses, which then guide a diffusion model to produce coherent and realistic dance video frames. Unlike traditional methods that primarily generate human motion in 3D, X-Dancer addresses data limitations and enhances scalability by modeling a wide spectrum of 2D dance motions, capturing their nuanced alignment with musical beats through readily available monocular videos. To achieve this, we first build a spatially compositional token representation from 2D human pose labels associated with keypoint confidences, encoding both large articulated body movements (e.g., upper and lower body) and fine-grained motions (e.g., head and hands). We then design a music-to-motion transformer model that autoregressively generates music-aligned dance pose token sequences, incorporating global attention to both musical style and prior motion context. Finally we leverage a diffusion backbone to animate the reference image with these synthesized pose tokens through AdaIN, forming a fully differentiable end-to-end framework. Experimental results demonstrate that X-Dancer is able to produce both diverse and characterized dance videos, substantially outperforming state-of-the-art methods in term of diversity, expressiveness and realism. Code and model will be available for research purposes.

Language diversity presents a significant challenge in speech-to-text (S2T) tasks, such as automatic speech recognition and translation. Traditional multi-task training approaches aim to address this by jointly optimizing multiple speech recognition and translation tasks across various languages. While models like Whisper, built on these strategies, demonstrate strong performance, they still face issues of high computational cost, language interference, suboptimal training configurations, and limited extensibility. To overcome these challenges, we introduce LoRS-Merging (low-rank and sparse model merging), a novel technique designed to efficiently integrate models trained on different languages or tasks while preserving performance and reducing computational overhead. LoRS-Merging combines low-rank and sparse pruning to retain essential structures while eliminating redundant parameters, mitigating language and task interference, and enhancing extensibility. Experimental results across a range of languages demonstrate that LoRS-Merging reduces the word error rate by 10% and improves BLEU scores by 4% compared to conventional multi-lingual multi-task training baselines. Our findings suggest that model merging, particularly LoRS-Merging, is a scalable and effective complement to traditional multi-lingual training strategies for S2T applications., Comment: 13 pages, submitted to ACL 2025

Tabular datasets are inherently heterogeneous, posing significant challenges for developing pre-trained foundation models. The recently introduced transformer-based Tabular Prior-data Fitted Network v2 (TabPFN v2) achieves unprecedented in-context learning accuracy across multiple tabular datasets, marking a pivotal advancement in tabular foundation models. In this paper, we comprehensively evaluate TabPFN v2 on over 300 datasets, confirming its exceptional generalization capabilities on small- to medium-scale tasks. Our analysis identifies randomized feature tokens as a key factor behind TabPFN v2's success, as they unify heterogeneous datasets into a fixed-dimensional representation, enabling more effective training and inference. To further understand TabPFN v2's predictions, we propose a leave-one-fold-out approach, transforming TabPFN v2 into a feature extractor and revealing its capability to simplify data distributions and boost accuracy. Lastly, to address TabPFN v2's limitations in high-dimensional, large-scale, and many-category tasks, we introduce a divide-and-conquer mechanism inspired by Chain-of-Thought prompting, enabling scalable inference. By uncovering the mechanisms behind TabPFN v2's success and introducing strategies to expand its applicability, this study provides key insights into the future of tabular foundation models.

As machine learning models become increasingly prevalent in time series applications, Explainable Artificial Intelligence (XAI) methods are essential for understanding their predictions. Within XAI, feature attribution methods aim to identify which input features contributed the most to a model's prediction, with their evaluation typically relying on perturbation-based metrics. Through empirical analysis across multiple datasets, model architectures, and perturbation strategies, we identify important class-dependent effects in these metrics: they show varying effectiveness across classes, achieving strong results for some while remaining less sensitive to others. In particular, we find that the most effective perturbation strategies often demonstrate the most pronounced class differences. Our analysis suggests that these effects arise from the learned biases of classifiers, indicating that perturbation-based evaluation may reflect specific model behaviors rather than intrinsic attribution quality. We propose an evaluation framework with a class-aware penalty term to help assess and account for these effects in evaluating feature attributions. Although our analysis focuses on time series classification, these class-dependent effects likely extend to other structured data domains where perturbation-based evaluation is common.

In 1958, Helson and Lowdenslager extended the theory of analytic functions to a general class of groups with ordered duals. In this context, analytic functions on such a group $G$ are defined as the integrable functions whose Fourier coefficients lie in the positive semigroup of the dual of $G$. In this paper, we found some applications of their theory to infinite-dimensional complex analysis. Specifically, we considered a special order on $\mathbb{Z}^\infty$ and corresponding analytic continuous functions on $\mathbb{T}^\omega$, which serves as the counterpart of the disk algebra in infinitely many variables setting. By characterizing its maximal ideals, we have generalized the following theorem to the infinite-dimensional case: For a positive function $w$ that is integrable and log-integrable on $\mathbb{T}^d$, there exists an outer function $g$ such that $w=|g|^2$ if and only if the support of $\hat{\log w}$ is a subset of $\mathbb{N}^d\cap (-\mathbb{N})^d$. Furthermore, we have found the counterpart of the above function algebra in the closed right half-plane, and the representing measures of each point in the right half-plane for this algebra. As an application of the order, we provided a new proof of the infinite-dimensional Szeg\"{o}'s theorem., Comment: 15pages, 0figures

In recent years, the rapid development of deep neural networks has brought increased attention to the security and robustness of these models. While existing adversarial attack algorithms have demonstrated success in improving adversarial transferability, their performance remains suboptimal due to a lack of consideration for the discrepancies between target and source models. To address this limitation, we propose a novel method, Inverse Knowledge Distillation (IKD), designed to enhance adversarial transferability effectively. IKD introduces a distillation-inspired loss function that seamlessly integrates with gradient-based attack methods, promoting diversity in attack gradients and mitigating overfitting to specific model architectures. By diversifying gradients, IKD enables the generation of adversarial samples with superior generalization capabilities across different models, significantly enhancing their effectiveness in black-box attack scenarios. Extensive experiments on the ImageNet dataset validate the effectiveness of our approach, demonstrating substantial improvements in the transferability and attack success rates of adversarial samples across a wide range of models.

Inspired by the BESIII observation of exotic hadron $G$(3900) [PRL 133(2024)081901] and by the study presented in a recent paper [PRL 133(2024)241903] where $G$(3900) was identified as a $P$-wave $D\bar D^*/\bar DD^*$ resonance, we use the parton and hadron cascade model PACIAE to simulate the $G$(3900) production in $e^+e^-$ annihilations at $\sqrt s$=4.95 GeV. The $G$(3900) candidates are recombined by Dynamically Constrained Phase-space Coalescence model using component mesons of $D\bar D$ or $D\bar D^*/\bar DD^*$ in the PACIAE simulated final hadronic state. We then calculate the $G$(3900) orbital angular momentum quantum number in its rest frame for the first time and perform the spectral classification for each of the above $G$(3900) candidates. Our results confirm the BESIII observation of $G$(3900) in the $e^+e^-\rightarrow D^0\bar D^0/ D^+D^-$ processes and suggest that the $G$(3900) composed of $D\bar D$ is $D$-wave dominant with $J^{PC}=2^{++}$. Moreover, our results indicate that the $G$(3900) composed of $D\bar D^*/\bar DD^*$ is $P$-wave dominant with $J^{PC}=$ $0^{-+}$, $1^{-+}$ or $2^{-+}$, which is consistent with the conclusion in PRL 133(2024)241903. Finally, significant discrepancies in the yields, the transverse momentum spectra and the rapidity distributions among the $G$(3900) $S$-wave, $P$-wave, and $D$-wave states are observed. These discrepancies are proposed as valuable criteria for deciphering the $G$(3900) orbital angular momentum quantum number and $J^{PC}$., Comment: five pages, 1 table, 2 figures

As a key infrastructure for China's "West-to-East Power Transmission" project, transmission lines (TL) face the threat of ice accretion under complex microclimatic conditions. This study proposes a plane principal strain sensing method based on a fiber Bragg grating circular array, achieving synchronous monitoring of 6 strains (ranging from -2000 to 2000 {\mu}{\epsilon}) across the TL cross-section. Through finite element simulation experiments, a mapping relationship between the bending of TL and the plane principal strain has been established. After completing the sensor calibration, an experimental platform for the freeze-thaw process of ice accretion on the TL was built. The relationships between ice mass and bending strain, as well as the ice shape on the TL cross-section (C-shaped and circular ice) and plane principal strain, were studied. Furthermore, a BP neural network model was developed to determine the 4 states of the icing process (no ice/freeze/stable/thaw), achieving an accuracy of 91.23%. This study provides effective monitoring of the freeze-thaw process of ice accretion on the TL, offering important technical support for the prevention and control of ice accretion in power grid.

Conformal Prediction is a widely studied technique to construct prediction sets of future observations. Most conformal prediction methods focus on achieving the necessary coverage guarantees, but do not provide formal guarantees on the size (volume) of the prediction sets. We first prove an impossibility of volume optimality where any distribution-free method can only find a trivial solution. We then introduce a new notion of volume optimality by restricting the prediction sets to belong to a set family (of finite VC-dimension), specifically a union of $k$-intervals. Our main contribution is an efficient distribution-free algorithm based on dynamic programming (DP) to find a union of $k$-intervals that is guaranteed for any distribution to have near-optimal volume among all unions of $k$-intervals satisfying the desired coverage property. By adopting the framework of distributional conformal prediction (Chernozhukov et al., 2021), the new DP based conformity score can also be applied to achieve approximate conditional coverage and conditional restricted volume optimality, as long as a reasonable estimator of the conditional CDF is available. While the theoretical results already establish volume-optimality guarantees, they are complemented by experiments that demonstrate that our method can significantly outperform existing methods in many settings., Comment: 41 pages, 19 figures, 2 tables

Large language models (LLMs) have significantly influenced various industries but suffer from a critical flaw, the potential sensitivity of generating harmful content, which poses severe societal risks. We developed and tested novel attack strategies on popular LLMs to expose their vulnerabilities in generating inappropriate content. These strategies, inspired by psychological phenomena such as the "Priming Effect", "Safe Attention Shift", and "Cognitive Dissonance", effectively attack the models' guarding mechanisms. Our experiments achieved an attack success rate (ASR) of 100% on various open-source models, including Meta's Llama-3.2, Google's Gemma-2, Mistral's Mistral-NeMo, Falcon's Falcon-mamba, Apple's DCLM, Microsoft's Phi3, and Qwen's Qwen2.5, among others. Similarly, for closed-source models such as OpenAI's GPT-4o, Google's Gemini-1.5, and Claude-3.5, we observed an ASR of at least 95% on the AdvBench dataset, which represents the current state-of-the-art. This study underscores the urgent need to reassess the use of generative models in critical applications to mitigate potential adverse societal impacts.