Search

Dataset license compliance is a critical yet complex aspect of developing commercial AI products, particularly with the increasing use of publicly available datasets. Ambiguities in dataset licenses pose significant legal risks, making it challenging even for software IP lawyers to accurately interpret rights and obligations. In this paper, we introduce LicenseGPT, a fine-tuned foundation model (FM) specifically designed for dataset license compliance analysis. We first evaluate existing legal FMs (i.e., FMs specialized in understanding and processing legal texts) and find that the best-performing model achieves a Prediction Agreement (PA) of only 43.75%. LicenseGPT, fine-tuned on a curated dataset of 500 licenses annotated by legal experts, significantly improves PA to 64.30%, outperforming both legal and general-purpose FMs. Through an A/B test and user study with software IP lawyers, we demonstrate that LicenseGPT reduces analysis time by 94.44%, from 108 seconds to 6 seconds per license, without compromising accuracy. Software IP lawyers perceive LicenseGPT as a valuable supplementary tool that enhances efficiency while acknowledging the need for human oversight in complex cases. Our work underscores the potential of specialized AI tools in legal practice and offers a publicly available resource for practitioners and researchers.

Low Earth Orbit (LEO) satellite networks have attracted considerable attention for their ability to deliver global, low-latency broadband Internet services. In this paper, we present a large-scale measurement study of the Starlink network, the largest LEO satellite constellation to date. We begin by proposing an efficient method for discovering active Starlink user routers, identifying approximately 3.2 million IPv6 addresses across 102 countries and 123 regions-representing, to the best of our knowledge, the most complete list of Starlink user routers' active IPv6 addresses. Based on the discovered user routers, we map the Starlink backbone network, which consists of 33 Points of Presence (PoPs) and 70 connections between them. Furthermore, we conduct a detailed statistical analysis of active Starlink users and PoPs. Finally, we summarize the IPv6 address assignment strategy adopted by the Starlink network. The dataset of the backbone network is publicly available at https://ki3.org.cn/#/starlink-network., Comment: 6 pages

Stellar activity contamination of radial velocity (RV) data is one of the top challenges plaguing the field of extreme precision RV (EPRV) science. Previous work has shown that photometry can be very effective at removing such signals from RV data, especially stellar activity caused by rotating star spots and plage.The exact utility of photometry for removing RV activity contamination, and the best way to apply it, is not well known. We present a combination photometric and RV study of eight Kepler/K2 FGK stars with known stellar variability. We use NEID RVs acquired simultaneously with TESS photometry, and we perform injection recovery tests to quantify the efficacy of recent TESS photometry versus archival Kepler/K2 photometry for removing stellar variability from RVs. We additionally experiment with different TESS sectors when training our models in order to quantify the real benefit of simultaneously acquired RVs and photometry. We conclude that Kepler photometry typically performs better than TESS at removing noise from RV data when it is available, likely due to longer baseline and precision. In contrast, for targets with available K2 photometry, especially those most active, and with high precision ($\sigma_{NEID}$ $<$ 1 m s$^{-1}$) NEID RVs, TESS may be the more informative dataset. However, contrary to expectations, we have found that training on simultaneous photometry does not always achieve the best results.

We present high resolution WIYN/NEID echelle spectroscopy (R $\approx70$,000) of the supernova (SN) 2023ixf in M101, obtained 1.51 to 18.51 days after explosion over nine epochs. Daily monitoring for the first four days after explosion shows narrow emission features ($\leq200$ km s$^{-1}$), exhibiting predominantly blueshifted velocities, that rapidly weaken, broaden, and vanish in a manner consistent with radiative acceleration and the SN shock eventually overrunning or enveloping the full extent of dense circumstellar medium (CSM). The most rapid evolution is in the He I emission, which is visible on day 1.51 but disappears by day 2.62. We measure the maximum pre-SN speed of He I to be 25 $^{+0}_{-5} \pm2$ km s$^{-1}$, where the error is attributable to the uncertainty in how much the He I had already been radiatively accelerated, and to measurement of the emission line profile. The radiative acceleration of material is likely driven by the shock-CSM interaction, and the CSM is accelerated to $\geq200$ km s$^{-1}$ before being completely swept up by the SN shock to $\sim 2000$ km s$^{-1}$. We compare the observed spectra with spherically-symmetric r16wb HERACLES/CMFGEN model spectra and find the line evolution to generally be consistent with radiative acceleration and optical depth effects. The progenitor of SN2023ixf underwent an enhanced mass loss phase $\gtrsim 4$ year prior to core-collapse, creating a dense, asymmetric CSM region extending out to approximately $r_{CSM} = 3.7 \times 10^{14}$ ($v_\textrm{shock}$/9500 km s$^{-1}$) cm., Comment: 20 pages, 13 figures, and 3 tables, Submitted to ApJ

Large language models (LLMs) are effective at capturing complex, valuable conceptual representations from textual data for a wide range of real-world applications. However, in fields like Intelligent Fault Diagnosis (IFD), incorporating additional sensor data-such as vibration signals, temperature readings, and operational metrics-is essential but it is challenging to capture such sensor data information within traditional text corpora. This study introduces a novel IFD approach by effectively adapting LLMs to numerical data inputs for identifying various machine faults from time-series sensor data. We propose FD-LLM, an LLM framework specifically designed for fault diagnosis by formulating the training of the LLM as a multi-class classification problem. We explore two methods for encoding vibration signals: the first method uses a string-based tokenization technique to encode vibration signals into text representations, while the second extracts statistical features from both the time and frequency domains as statistical summaries of each signal. We assess the fault diagnosis capabilities of four open-sourced LLMs based on the FD-LLM framework, and evaluate the models' adaptability and generalizability under various operational conditions and machine components, namely for traditional fault diagnosis, cross-operational conditions, and cross-machine component settings. Our results show that LLMs such as Llama3 and Llama3-instruct demonstrate strong fault detection capabilities and significant adaptability across different operational conditions, outperforming state-of-the-art deep learning (DL) approaches in many cases., Comment: 20 pages, 2 figures, 16 tables, including the tables in the appendix

Br\"{a}nd\'{e}n and Claesson introduced the concept of mesh patterns in 2011, and since then, these patterns have attracted significant attention in the literature. Subsequently, in 2015, Hilmarsson \emph{et al.} initiated the first systematic study of avoidance of mesh patterns, while Kitaev and Zhang conducted the first systematic study of the distribution of mesh patterns in 2019. A permutation $\sigma = \sigma_1 \sigma_2 \cdots \sigma_n$ in the symmetric group $S_n$ is called a king permutation if $\left| \sigma_{i+1}-\sigma_i \right| > 1$ for each $1 \leq i \leq n-1$. Riordan derived a recurrence relation for the number of such permutations in 1965. The generating function for king permutations was obtained by Flajolet and Sedgewick in 2009. In this paper, we initiate a systematic study of the distribution of mesh patterns on king permutations by finding distributions for 22 mesh patterns of short length., Comment: 26 pages

Water molecules predominantly act as solvents in electrochemical systems and are often modeled as a passive dielectric medium. In this work, we use molecular dynamics simulations and theoretical analysis to revisit this conventional view. We reveal that the interfacial polarized water overscreens the electrostatic potential between ions and the surface beyond being a passive dielectric medium. This overscreening enables the interfacial water to dominate the electric potential spatial distribution, inverting the electrode surface potential polarity and dominating the capacitance. A model is then developed to incorporate this critical interfacial water polarization.

The significant inter-subject variability in electroencephalogram (EEG) signals often leads to knowledge being overwritten as new tasks are introduced in continual EEG decoding. While retraining on the entire dataset with each new input can prevent forgetting, this approach incurs high computational costs. An ideal brain-computer interface (BCI) model should continuously learn new information without retraining from scratch, thus reducing these costs. Most transfer learning models rely on large source-domain datasets for pre-training, yet data availability is frequently limited in real-world applications due to privacy concerns. Furthermore, such models are prone to catastrophic forgetting in continual EEG decoding tasks. To address these challenges, we propose a personalized subject-incremental learning (SIL) framework for continual EEG decoding that integrates Euclidean Alignment for fast domain adaptation, an exemplar replay mechanism to retain prior knowledge, and reservoir sampling-based memory management to handle memory constraints in long-term learning. Validated on the OpenBMI dataset with 54 subjects, our framework effectively balances knowledge retention with classification performance in continual MI-EEG tasks, offering a scalable solution for real-world BCI applications.

For $F\subseteq V(G)$, if $G-F$ is a disconnected graph with at least $r$ components and each vertex $v\in V(G)\backslash F$ has at least $g$ neighbors, then $F$ is called a $g$-good $r$-component cut of $G$. The $g$-good $r$-component connectivity of $G$, denoted by $c\kappa_{g,r}(G)$, is the minimum cardinality of $g$-good $r$-component cuts of $G$. Let $\mathcal{G}_n^{k,\delta}$ be the set of graphs of order $n$ with minimum degree $\delta$ and $g$-good $r$-component connectivity $c\kappa_{g,r}(G)=k$. In the paper, we determine the extremal graphs attaining the maximum spectral radii among all graphs in $\mathcal{G}_n^{k,\delta}$. A subset $F\subseteq V(G)$ is called a $g$-good neighbor cut of $G$ if $G-F$ is disconnected and each vertex $v\in V(G)\backslash F$ has at least $g$ neighbors. The $g$-good neighbor connectivity $\kappa_g(G)$ of a graph $G$ is the minimum cardinality of $g$-good neighbor cuts of $G$. The condition of $g$-good neighbor connectivity is weaker than that of $g$-good $r$-component connectivity, and there is no requirement on the number of components. As a counterpart, we also study similar problem for $g$-good neighbor connectivity.

The attached-eddy model (AEM) predicts mean velocity and streamwise velocity variance profiles that follow a logarithmic shape in the overlap region of high Reynolds number wall-bounded turbulent flows. Moreover, the AEM coefficients are presumed to attain asymptotically constant values at very high Reynolds numbers. Here, the logarithmic behaviour of the AEM predictions in the near-neutral atmospheric surface layer is examined using sonic anemometer measurements from a 62-m meteorological tower located in the Eastern Snake River Plain, Idaho, US. Utilizing an extensive 210-day dataset, the inertial sublayer (ISL) is first identified by analyzing the measured momentum flux and mean velocity profile. The logarithmic behaviour of the streamwise velocity variance and the associated `-1' scaling of the streamwise velocity energy spectra are then investigated. The findings indicate that the Townsend-Perry coefficient ($A_1$) is influenced by mild non-stationarity that manifests itself as a Reynolds number dependence. After excluding non-stationary runs and requiring a Reynolds number higher than $4 \times 10^7$, the inferred $A_1$ converges to values ranging between 1 and 1.25, consistent with laboratory experiments. Moreover, the independence of the normalized vertical velocity variance from the wall-normal distance in the ISL is further checked and the constant coefficient value agrees with reported laboratory experiments at very high Reynolds numbers as well as many surface layer experiments. Furthermore, nine benchmark cases selected through a restrictive quality control reveal a closer relationship between the `-1' scaling in the streamwise velocity energy spectrum and the logarithmic behaviour of streamwise velocity variance at higher Reynolds numbers, though no direct equivalence between them is observed.

Multimodal learning has been a popular area of research, yet integrating electroencephalogram (EEG) data poses unique challenges due to its inherent variability and limited availability. In this paper, we introduce a novel multimodal framework that accommodates not only conventional modalities such as video, images, and audio, but also incorporates EEG data. Our framework is designed to flexibly handle varying input sizes, while dynamically adjusting attention to account for feature importance across modalities. We evaluate our approach on a recently introduced emotion recognition dataset that combines data from three modalities, making it an ideal testbed for multimodal learning. The experimental results provide a benchmark for the dataset and demonstrate the effectiveness of the proposed framework. This work highlights the potential of integrating EEG into multimodal systems, paving the way for more robust and comprehensive applications in emotion recognition and beyond.

The inner shadow of a black hole, as a projection of the event horizon, is regarded as a potential tool for testing gravitational theories and constraining system parameters. Whether this holds in the case of a tilted accretion disk warrants further investigation. In this paper, we employ a ray-tracing algorithm to simulate images of the Kerr black hole illuminated by a tilted thin accretion disk, with particular attention to the relationship between the inner shadow and system parameters. Our findings reveal that in the case of an equatorial accretion disk, the Kerr black hole exhibits a minimum inner shadow size of $S_{\textrm{min}} = 13.075$ M$^{2}$, where M denotes the black hole mass. This minimum is achieved when the viewing angle is $0^{\circ}$ and the spin parameter approaches $1$. However, with a non-zero disk tilt, the inner shadow exhibits novel configurations--taking on petal, crescent, or eyebrow shapes--significantly smaller than $S_{\textrm{min}}$ across various parameter spaces. This indicates that the inner shadow is highly sensitive to the accretion environment, suggesting that caution is needed when using it as a diagnostic tool for black holes. Notably, an observed inner shadow smaller than $S_{\textrm{min}}$ would either indicate the presence of a tilted accretion disk or support the viability of modified gravity. Moreover, in certain parameter spaces, we identify the emergence of a dual-shadow structure, which could also serve as a probe for the tilted accretion disk., Comment: 23 pages, 8 figures, references added

Unlike images and natural language tokens, time series data is highly semantically sparse, resulting in labor-intensive label annotations. Unsupervised and Semi-supervised Domain Adaptation (UDA and SSDA) have demonstrated efficiency in addressing this issue by utilizing pre-labeled source data to train on unlabeled or partially labeled target data. However, in domain adaptation methods designed for downstream classification tasks, directly adapting labeled source samples with unlabelled target samples often results in similar distributions across various classes, thereby compromising the performance of the target classification task. To tackle this challenge, we proposed a Global-Local Alignment Domain Adaptation (GLA-DA) method for multivariate time series data. Data from two domains were initially encoded to align in an intermediate feature space adversarially, achieving Global Feature Alignment (GFA). Subsequently, GLA-DA leveraged the consistency between similarity-based and deep learning-based models to assign pseudo labels to unlabeled target data. This process aims to preserve differences among data with distinct labels by aligning the samples with the same class labels together, achieving Local Class Alignment (LCA). We implemented GLA-DA in both UDA and SSDA scenarios, showcasing its superiority over state-of-the-art methods through extensive experiments on various public datasets. Ablation experiments underscored the significance of key components within GLA-DA.

Let $\emph{spex}_{\mathcal{OP}}(n,F)$ and $\emph{spex}_{\mathcal{P}}(n,F)$ be the maximum spectral radius over all $n$-vertex $F$-free outerplanar graphs and planar graphs, respectively. Define $tC_l$ as $t$ vertex-disjoint $l$-cycles, $B_{tl}$ as the graph obtained by sharing a common vertex among $t$ edge-disjoint $l$-cycles %$B_{tl}$ as the graph obtained by connecting all cycles in $tC_l$ at a single vertex, and $(t+1)K_{2}$ as the disjoint union of $t+1$ copies of $K_2$. In the 1990s, Cvetkovi\'c and Rowlinson conjectured $K_1 \vee P_{n-1}$ maximizes spectral radius in outerplanar graphs on $n$ vertices, while Boots and Royle (independently, Cao and Vince) conjectured $K_2 \vee P_{n-2} $ does so in planar graphs. Tait and Tobin [J. Combin. Theory Ser. B, 2017] determined the fundamental structure as the key to confirming these two conjectures for sufficiently large $n.$ Recently, Fang et al. [J. Graph Theory, 2024] characterized the extremal graph with $\emph{spex}_{\mathcal{P}}(n,tC_l)$ in planar graphs by using this key. In this paper, we first focus on outerplanar graphs and adopt a similar approach to describe the key structure of the connected extremal graph with $\emph{spex}_{\mathcal{OP}}(n,F)$, where $F$ is contained in $K_1 \vee P_{n-1}$ but not in $K_{1} \vee ((t-1)K_2\cup(n-2t+1)K_1)$. Based on this structure, we determine $\emph{spex}_{\mathcal{OP}}(n,B_{tl})$ and $\emph{spex}_{\mathcal{OP}}(n,(t+1)K_{2})$ along with their unique extremal graphs for all $t\geq1$, $l\geq3$ and large $n$. Moreover, we further extend the results to planar graphs, characterizing the unique extremal graph with $\emph{spex}_{\mathcal{P}}(n,B_{tl})$ for all $t\geq3$, $l\geq3$ and large $n$., Comment: arXiv admin note: text overlap with arXiv:2304.06942 by other authors

WebAssembly (Wasm) is a low-level bytecode format that can run in modern browsers. With the development of standalone runtimes and the improvement of the WebAssembly System Interface (WASI), Wasm has further provided a more complete sandboxed runtime experience for server-side applications, effectively expanding its application scenarios. However, the implementation of WASI varies across different runtimes, and suboptimal interface implementations can lead to performance degradation during interactions between the runtime and the operating system. Existing research mainly focuses on overall performance evaluation of runtimes, while studies on WASI implementations are relatively scarce. To tackle this problem, we propose an eBPF-based WASI performance analysis framework. It collects key performance metrics of the runtime under different I/O load conditions, such as total execution time, startup time, WASI execution time, and syscall time. We can comprehensively analyze the performance of the runtime's I/O interactions with the operating system. Additionally, we provide a detailed analysis of the causes behind two specific WASI performance anomalies. These analytical results will guide the optimization of standalone runtimes and WASI implementations, enhancing their efficiency.

AllReduce is an important and popular collective communication primitive, which has been widely used in areas such as distributed machine learning and high performance computing. To design, analyze, and choose from various algorithms and implementations of AllReduce, the time cost model plays a crucial role, and the predominant one is the $(\alpha,\beta,\gamma)$ model. In this paper, we revisit this model, and reveal that it cannot well characterize the time cost of AllReduce on modern clusters; thus must be updated. We perform extensive measurements to identify two additional terms contributing to the time cost: the incast term and the memory access term. We augment the $(\alpha,\beta,\gamma)$ model with these two terms, and present GenModel as a result. Using GenModel, we discover two new optimalities for AllReduce algorithms, and prove that they cannot be achieved simultaneously. Finally, striking the balance between the two new optimalities, we design GenTree, an AllReduce plan generation algorithm specialized for tree-like topologies. Experiments on a real testbed with 64 GPUs show that GenTree can achieve 1.22$\times$ to 1.65$\times$ speed-up against NCCL. Large-scale simulations also confirm that GenTree can improve the state-of-the-art AllReduce algorithm by a factor of $1.2$ to $7.4$ in scenarios where the two new terms dominate.

The projection effect of an anisotropic accretion disk causes its electromagnetic radiation to depend on the emission angle. Although this dependency has the potential to influence the observational characteristics of black holes, it has not received sufficient attention. In this paper, we employ a relativistic ray-tracing algorithm to numerically simulate 86 GHz and 230 GHz images of deformed Schwarzschild black holes illuminated by an equatorial anisotropic accretion disk, aiming to reveal the observational signatures of the target black hole and the impact of the projection effect on the images. The study demonstrates that while the introduction of the projection effect does not alter the profiles of the black hole's inner shadow and critical curve, it significantly suppresses the specific intensity of light rays, particularly in direct emission, thereby reducing image brightness. The extent of this reduction depends on both the observation inclination and frequency. This phenomenon aids in the extraction of geometric information from higher-order subrings in the image. Furthermore, we find that increasing the deformation parameter enhances the brightness of the deformed Schwarzschild black hole image, accompanied by a reduction in the size of the critical curve and inner shadow. This relationship establishes a connection between the intrinsic properties of deformed Schwarzschild black holes and their observational characteristics, providing a reliable tool for testing the no-hair theorem and gravitational theories. Specifically, we propose a novel method for constraining parameters based on the silhouette of the inner shadow, which holds promise for extension to any spherically symmetric black hole in other gravitational theories., Comment: 30 pages, 14 figures

Vulnerabilities related to option combinations pose a significant challenge in software security testing due to their vast search space. Previous research primarily addressed this challenge through mutation or filtering techniques, which inefficiently treated all option combinations as having equal potential for vulnerabilities, thus wasting considerable time on non-vulnerable targets and resulting in low testing efficiency. In this paper, we utilize carefully designed prompt engineering to drive the large language model (LLM) to predict high-risk option combinations (i.e., more likely to contain vulnerabilities) and perform fuzz testing automatically without human intervention. We developed a tool called ProphetFuzz and evaluated it on a dataset comprising 52 programs collected from three related studies. The entire experiment consumed 10.44 CPU years. ProphetFuzz successfully predicted 1748 high-risk option combinations at an average cost of only \$8.69 per program. Results show that after 72 hours of fuzzing, ProphetFuzz discovered 364 unique vulnerabilities associated with 12.30\% of the predicted high-risk option combinations, which was 32.85\% higher than that found by state-of-the-art in the same timeframe. Additionally, using ProphetFuzz, we conducted persistent fuzzing on the latest versions of these programs, uncovering 140 vulnerabilities, with 93 confirmed by developers and 21 awarded CVE numbers., Comment: Preprint

We present a new analysis of Kepler-21, the brightest (V = 8.5) Kepler system with a known transiting exoplanet, Kepler-21 b. Kepler-21 b is a radius valley planet ($R = 1.6\pm 0.2 R_{\oplus}$) with an Earth-like composition (8.38$\pm$1.62 g/cc), though its mass and radius fall in the regime of possible "water worlds." We utilize new Keck/HIRES and WIYN/NEID radial velocity (RV) data in conjunction with Kepler and TESS photometry to perform a detailed study of activity mitigation between photometry and RVs. We additionally refine the system parameters, and we utilize Gaia astrometry to place constraints on a long-term RV trend. Our activity analysis affirms the quality of Kepler photometry for removing correlated noise from RVs, despite its temporal distance, though we reveal some cases where TESS may be superior. Using refined orbital parameters and updated composition curves, we rule out a ``water world" scenario for Kepler-21 b, and we identify a long period super-Jupiter planetary candidate, Kepler-21 (c).

The WIYN 3.5m Telescope at Kitt Peak National Observatory hosts a suite of optical and near infrared instruments, including an extreme precision, optical spectrograph, NEID, built for exoplanet radial velocity studies. In order to achieve sub ms precision, NEID has strict requirements on survey efficiency, stellar image positioning, and guiding performance, which have exceeded the native capabilities of the telescope's original pointing and tracking system. In order to improve the operational efficiency of the telescope we have developed a novel telescope pointing system, built on a recurrent neural network, that does not rely on the usual pointing models (TPoint or other quasi physical bases). We discuss the development of this system, how the intrinsic properties of the pointing problem inform our network design, and show preliminary results from our best models. We also discuss plans for the generalization of this framework, so that it can be applied at other sites.

A graph is trivial if it contains one vertex and no edges. The essential connectivity $\kappa^{\prime}$ of $G$ is defined to be the minimum number of vertices of $G$ whose removal produces a disconnected graph with at least two non-trivial components. Let $\mathcal{A}_n^{\kappa',\delta}$ be the set of graphs of order $n$ with minimum degree $\delta$ and essential connectivity $\kappa'$. In this paper, we determine the graphs attaining the maximum spectral radii among all graphs in $\mathcal{A}_n^{\kappa',\delta}$ and characterize the corresponding extremal graphs. In addition, we also determine the digraphs which achieve the maximum spectral radii among all strongly connected digraphs with given essential connectivity and give the exact values of the spectral radii of these digraphs.

RAG systems face limitations when semantic relevance alone does not guarantee improved generation quality. This issue becomes particularly evident due to the sensitivity of large language models (LLMs) to the ordering of few-shot prompts, which can affect model performance. To address this challenge, aligning LLM outputs with human preferences using structured feedback, such as options to copy, regenerate, or dislike, offers a promising method for improvement. This feedback is applied to the entire list of inputs rather than giving specific ratings for individual documents, making it a Listwide Labels Learning-to-Rank task. To address this task, we propose Pistis-RAG, a new RAG framework designed with a content-centric approach to better align LLMs with human preferences. Pistis-RAG effectively utilizes human feedback, enhancing content ranking and generation quality. To validate our framework, we use public datasets to simulate human feedback, allowing us to evaluate and refine our method effectively. Experimental results indicate that Pistis-RAG improves alignment with human preferences relative to the baseline RAG system, showing a 6.06% increase in MMLU (English) and a 7.08% increase in C-EVAL (Chinese) accuracy metrics. These results highlight Pistis-RAG's effectiveness in overcoming the limitations associated with traditional RAG approaches.

Limited by the scale and diversity of time series data, the neural networks trained on time series data often overfit and show unsatisfacotry performances. In comparison, large language models (LLMs) recently exhibit impressive generalization in diverse fields. Although massive LLM based approaches are proposed for time series tasks, these methods require to load the whole LLM in both training and reference. This high computational demands limit practical applications in resource-constrained settings, like edge-computing and IoT devices. To address this issue, we propose Knowledge Pruning (KP), a novel paradigm for time series learning in this paper. For a specific downstream task, we argue that the world knowledge learned by LLMs is much redundant and only the related knowledge termed as "pertinent knowledge" is useful. Unlike other methods, our KP targets to prune the redundant knowledge and only distill the pertinent knowledge into the target model. This reduces model size and computational costs significantly. Additionally, different from existing LLM based approaches, our KP does not require to load the LLM in the process of training and testing, further easing computational burdens. With our proposed KP, a lightweight network can effectively learn the pertinent knowledge, achieving satisfactory performances with a low computation cost. To verify the effectiveness of our KP, two fundamental tasks on edge-computing devices are investigated in our experiments, where eight diverse environments or benchmarks with different networks are used to verify the generalization of our KP. Through experiments, our KP demonstrates effective learning of pertinent knowledge, achieving notable performance improvements in regression (19.7% on average) and classification (up to 13.7%) tasks, showcasing state-of-the-art results., Comment: 12 pages, 5 figures

Few-shot named entity recognition (NER) systems recognize entities using a few labeled training examples. The general pipeline consists of a span detector to identify entity spans in text and an entity-type classifier to assign types to entities. Current span detectors rely on extensive manual labeling to guide training. Almost every span detector requires initial training on basic span features followed by adaptation to task-specific features. This process leads to repetitive training of the basic span features among span detectors. Additionally, metric-based entity-type classifiers, such as prototypical networks, typically employ a specific metric that gauges the distance between the query sample and entity-type referents, ultimately assigning the most probable entity type to the query sample. However, these classifiers encounter the sample dependency problem, primarily stemming from the limited samples available for each entity-type referent. To address these challenges, we proposed an improved few-shot NER pipeline. First, we introduce a steppingstone span detector that is pre-trained on open-domain Wikipedia data. It can be used to initialize the pipeline span detector to reduce the repetitive training of basic features. Second, we leverage a large language model (LLM) to set reliable entity-type referents, eliminating reliance on few-shot samples of each type. Our model exhibits superior performance with fewer training steps and human-labeled data compared with baselines, as demonstrated through extensive experiments on various datasets. Particularly in fine-grained few-shot NER settings, our model outperforms strong baselines, including ChatGPT. We will publicly release the code, datasets, LLM outputs, and model checkpoints., Comment: ieee access: https://doi.org/10.1109/ACCESS.2024.3374727

To control structural responses under various actions, the growing use of supplementary damping systems in modern civil engineering structures necessitates inspecting and evaluating their operational performance postinstallation. However, due to the dispersed placement and complex nonlinearities of these devices, difficulties arise in determining minimal sensor configuration. This is inherently connected to a pivotal challenge: establishing a reliable input-output mapping, which comprises both the mathematical model and sensor arrangements. Prior work indicates this can be achieved through theoretical observability analysis or Lie symmetries analysis, both of which provide different perspectives on the existence of a way to access the solutions of a system identification problem uniquely (at least locally). The present study introduces a unified framework, enhanced by algorithm realization as an application guide, for analyzing the observability and Lie symmetries of a given input-output mapping. We demonstrate its implementation via examples of a building structure with various damping systems under different conditions such as seismic loads, wind loads, and operational vibrations. Finally, we present a case study for an isolation building with an inerter damper and minimal sensor arrangement under seismic action. The results demonstrate that the unscented Kalman filter, a system identification method, can precisely estimate structural responses and assess damping device performance once a reliable input-output mapping is established.

For two integers $r\geq 2$ and $h\geq 0$, the \emph{$h$-extra $r$-component connectivity} $\kappa^h_r(G)$ of a graph $G$ is defined to be the minimum size of a subset of vertices whose removal disconnects $G$, and there are at least $r$ connected components in $G\!-\!S$ and each component has at least $h+1$ vertices. Denote by $\mathcal{G}_{n,\delta}^{\kappa_r^h}$ the set of graphs with $h$-extra $r$-component connectivity $\kappa^h_r(G)$ and minimum degree $\delta$. The following problem concerning spectral radius was proposed by Brualdi and Solheid [On the spectral radius of complementary acyclic matrices of zeros and one, SIAM J. Algebra Discrete Methods 7 (1986) 265-272]: Given a set of graphs $\mathscr{S}$, find an upper bound for the spectral radius of graphs in $\mathscr{S}$ and characterize the graphs in which the maximal spectral radius is attained. We study this question for $\mathscr{S}=\mathcal{G}_{n,\delta}^{\kappa_r^h}$ where $r\geq 2$ and $h\geq 0$. Fan, Gu and Lin [$l$-connectivity, $l$-edge-connectivity and spectral radius of graphs, \emph{arXiv}:2309.05247] give the answer to $r\geq 2$ and $h=0$. In this paper, we solve this problem completely for $r\geq 2$ and $h\geq1$. Moreover, we also investigate analogous problems for the edge version. Our results can break the restriction of the extremum structure of the conditional connectivity. This implies some previous results in connectivity and edge-connectivity.

Blockchain technology and Artificial Intelligence (AI) have emerged as transformative forces in their respective domains. This paper explores synergies and challenges between these two technologies. Our research analyses the biggest projects combining blockchain and AI, based on market capitalization, and derives a novel framework to categorize contemporary and future use cases. Despite the theoretical compatibility, current real-world applications combining blockchain and AI remain in their infancy.

A signed graph $\Sigma=(G,\sigma)$ consists of an underlying graph $G=(V,E)$ with a sign function $\sigma:E\rightarrow\{-1,1\}$. Let $A(\Sigma)$ be the adjacency matrix of $\Sigma$ and $\lambda_1(\Sigma)$ denote the largest eigenvalue (index) of $\Sigma$.Define $(K_n,H^-)$ as a signed complete graph whose negative edges induce a subgraph $H$. In this paper, we focus on the following problem: which spanning tree $T$ with a given number of pendant vertices makes the $\lambda_1(A(\Sigma))$ of the unbalanced $(K_n,T^-)$ as large as possible? To answer the problem, we characterize the extremal signed graph with maximum $\lambda_1(A(\Sigma))$ among graphs of type $(K_n,T^-)$.

Optical skyrmions are an emerging class of structured light with sophisticated particle-like topologies with great potential for revolutionizing modern informatics. However, the current generation of optical skyrmions involves complex or bulky systems, hindering their development of practical applications. Here, exploiting the emergent "lab-on-fiber" technology, we demonstrate the design of a metafiber-integrated photonic skyrmion generator. We not only successfully generated high-quality optical skyrmions from metafibers, but also experimentally verified their remarkable properties, such as regulability and topological stability with deep-subwavelength features beyond the diffraction limits. Our flexible and fiber-integrated optical skyrmions platform paves the avenue for future applications of topologically-enhanced remote super-resolution microscopy and super-robust information transfer.