Search

Multi-object tracking is advancing through two dominant paradigms: traditional tracking by detection and newly emerging tracking by query. In this work, we fuse them together and propose the tracking-by-detection-and-query paradigm, which is achieved by a Learnable Associator. Specifically, the basic information interaction module and the content-position alignment module are proposed for thorough information Interaction among object queries. Tracking results are directly Decoded from these queries. Hence, we name the method as LAID. Compared to tracking-by-query models, LAID achieves competitive tracking accuracy with notably higher training efficiency. With regard to tracking-by-detection methods, experimental results on DanceTrack show that LAID significantly surpasses the state-of-the-art heuristic method by 3.9% on HOTA metric and 6.1% on IDF1 metric. On SportsMOT, LAID also achieves the best score on HOTA metric. By holding low training cost, strong tracking capabilities, and an elegant end-to-end approach all at once, LAID presents a forward-looking direction for the field.

Utilizing a range of techniques including multi-band light curves, softness ratio analysis, structure functions, rms spectra, cross-correlation functions, and ratios of spectra from different intervals, we present a comprehensive study of the complex X-ray spectral variability in Seyfert 1 galaxy Ark 120, through re-analyzing its six XMM-Newton observations taken between 2003 and 2014. We find a clear ''softer-when-brighter" trend in the 2--10 keV power-law component over long timescales, with this trend being timescale dependent, as it is much weaker on shorter timescales, similar to that previously detected in NGC 4051. Notably, a rare ''harder-when-brighter" trend is observed during one exposure, indicating dynamic changes in the spectral variability behavior of the power-law component. This exceptional exposure, with the spectral variability indeed marked by a power-law pivoting at an unusually low energy of ~ 2 keV, suggests intricate variations in the thermal Comptonization processes within the corona. Furthermore, when the data below 2 keV are included, we identify that the soft excess component adds significant complexity to the spectral variability, such as evidenced by a transition from ''harder-when-brighter'' to ''softer-when-brighter'' during another single exposure. Such extra complexity arises because the variability of the soft excess sometimes follows and sometimes does not follow the changes in the power-law component. Our findings underscore the necessity of applying multiple analytic techniques to fully capture the multifaceted spectral variability of AGNs., Comment: 13 pages, 11 figures, submitted. Comments are very welcome!

We focus on two hypergraph Ramsey problems. First, we consider the Erd\H{o}s-Hajnal function $r_k(k+1,t;n)$. In 1972, Erd\H{o}s and Hajnal conjectured that the tower growth rate of $r_k(k+1,t;n)$ is $t-1$ for each $2\le t\le k$. To finish this conjecture, it remains to show that the tower growth rate of $r_4(5,4;n)$ is three. We prove a superexponential lower bound for $r_4(5,4;n)$, which improves the previous best lower bound $r_4(5,4;n)\geq 2^{\Omega(n^2)}$ from Mubayi and Suk (\emph{J. Eur. Math. Soc., 2020}). Second, we prove an upper bound for the hypergraph Erd\H{o}s-Rogers function $f^{(k)}_{k+1,k+2}(N)$ that is an iterated $(k-3)$-fold logarithm in $N$ for each $k\geq 5$. This improves the previous upper bound that is an iterated $(k-13)$-fold logarithm in $N$ for $k\ge14$ due to Mubayi and Suk (\emph{J. London Math. Soc., 2018}), in which they conjectured that $f^{(k)}_{k+1,k+2}(N)$ is an iterated $(k-2)$-fold logarithm in $N$ for each $k\ge3$., Comment: 18 pages

As an emerging magnetic phase, altermagnets with compensated magnetic order and non-relativistic spin-splitting have attracted widespread attention. Currently, strain engineering is considered to be an effective method for inducing valley polarization in altermagnets, however, achieving controllable switching of valley polarization is extremely challenging. Herein, combined with tight-binding model and first-principles calculations, we propose that interlayer sliding can be used to successfully induce and effectively manipulate the large valley polarization in altermagnets. Using Fe2MX4 (M = Mo, W; X = S, Se or Te) family as examples, we predict that sliding induced ferrovalley states in such systems can exhibit many unique properties, including the linearly optical dichroism that is independent of spin-orbit coupling, and the anomalous valley Hall effect. These findings imply the correlation among spin, valley, layer and optical degrees of freedom that makes altermagnets attractive in spintronics, valleytronics and even their crossing areas.

Large language models (LLMs) enhanced with retrieval-augmented generation (RAG) have introduced a new paradigm for web search. However, the limited context awareness of LLMs degrades their performance on RAG tasks. Existing methods to enhance context awareness are often inefficient, incurring time or memory overhead during inference, and many are tailored to specific position embeddings. In this paper, we propose Position-Embedding-Agnostic attention Re-weighting (PEAR), which enhances the context awareness of LLMs with zero inference overhead. Specifically, on a proxy task focused on context copying, we first detect heads which suppress the models' context awareness thereby diminishing RAG performance. To weaken the impact of these heads, we re-weight their outputs with learnable coefficients. The LLM (with frozen parameters) is optimized by adjusting these coefficients to minimize loss on the proxy task. As a result, the coefficients are optimized to values less than one, thereby reducing their tendency to suppress RAG performance. During inference, the optimized coefficients are fixed to re-weight these heads, regardless of the specific task at hand. Our proposed PEAR offers two major advantages over previous approaches: (1) It introduces zero additional inference overhead in terms of memory usage or inference time, while outperforming competitive baselines in accuracy and efficiency across various RAG tasks. (2) It is independent of position embedding algorithms, ensuring broader applicability., Comment: preprint

Approximation of problems in linear elasticity having small shear modulus in a thin region is considered. Problems of this type arise when modeling ground motion due to earthquakes where rupture occurs in a thin fault. It is shown that, under appropriate scaling, solutions of these problems can be approximated by solutions of a limit problem where the fault region is represented by a surface. In a numerical context this eliminates the need to resolve the large deformations in the fault; a numerical example is presented to illustrate efficacy of this strategy.

Recently, a flurry of works have found strong competition between charge density wave (CDW) and superconductivity (SC) in the doped Hubbard and $t$-$J$ models on the square lattice. Interestingly, some recent results suggest that the electron-phonon coupling may suppress CDW order and enhance SC. In this work, we consider the square-lattice Hubbard model with the Holstein or Su-Schrieffer-Heeger electron-phonon coupling at the large-$U$ and antiadiabatic (infinite phonon frequency) limit, which gives an effective $t$-$J$ model with either a density attractive interaction $V$ or a $J_P$ term that contributes a larger spin exchange and a density repulsive interaction. To explore how these effective couplings may suppress CDW and give a SC, we implement the density matrix renormalization group simulation on the $t$-$t'$-$J$ model with $V$ or $J_P$ coupling. We focus on the {\it six-leg} cylinder system with the next-nearest-neighbor hopping $t'$, which hosts partially filled stripe and $d$-wave SC in phase diagram. By tuning $t'/t > 0$ and $V$ or $J_P$, we establish two quantum phase diagrams. In the SC phases, the increased $V$ or $J_P$ coupling can enhance the quasi-long-range SC order, consistent with some previous findings. Nonetheless, no SC emerges when the partially filled stripe phase disappears with increased $V$ or $J_P$. Instead, the system has a transition to either a phase-separation-like regime or a filled stripe phase. On the other hand, with increased $t'/t$, not only the partially filled stripe but the phase separation and filled stripe can also be tuned to SC phase. Our results suggest that although $V$ and $J_P$ couplings may strengthen hole binding, the hole dynamics controlled by $t'/t$ appears to play more crucial role for obtaining a SC in $t$-$J$ model., Comment: 12 pages, 10 figures

The discovery of single photon emitters (SPEs) in two-dimensional (2D) layered materials has greatly inspired numerous studies towards utilizing the system for quantum science and technology. Thus, the dynamic control of SPEs, including neutral and charged emitters, is highly desirable. In addition to the electric control, strain tuning is particularly attractive for the 2D materials since it can activate SPEs which are formed upon localizing free excitons. While strain engineering has been demonstrated for free and neutral localized excitons, few were shown on charged localized excitons which require an additional gate control. In this article, we show the strain-tunable charged localized excitons by transferring a top-gated monolayer semiconductor on a relaxor ferroelectric. Importantly, we unveil an enhanced interaction between the localized oscillating dipoles and the nanodomains. We further demonstrate the strain-dependent circular polarization and tunable rates of energy shifts under a magnetic field. Our results imply that the integration of 2D materials with relaxor ferroelectrics provides a rich platform for nanophotonics and quantum photonics., Comment: 11 pages, 4 figures

Identifying structures in common forms the basis for networked systems design and optimization. However, real structures represented by graphs are often of varying sizes, leading to the low accuracy of traditional graph classification methods. These graphs are called cross-scale graphs. To overcome this limitation, in this study, we propose GSpect, an advanced spectral graph filtering model for cross-scale graph classification tasks. Compared with other methods, we use graph wavelet neural networks for the convolution layer of the model, which aggregates multi-scale messages to generate graph representations. We design a spectral-pooling layer which aggregates nodes to one node to reduce the cross-scale graphs to the same size. We collect and construct the cross-scale benchmark data set, MSG (Multi Scale Graphs). Experiments reveal that, on open data sets, GSpect improves the performance of classification accuracy by 1.62% on average, and for a maximum of 3.33% on PROTEINS. On MSG, GSpect improves the performance of classification accuracy by 15.55% on average. GSpect fills the gap in cross-scale graph classification studies and has potential to provide assistance in application research like diagnosis of brain disease by predicting the brain network's label and developing new drugs with molecular structures learned from their counterparts in other systems.

Randomized experiments are the gold standard for estimating treatment effects, yet network interference challenges the validity of traditional estimators by violating the stable unit treatment value assumption and introducing bias. While cluster randomized experiments mitigate this bias, they encounter limitations in handling network complexity and fail to distinguish between direct and indirect effects. To address these challenges, we develop a design-based asymptotic theory for the existing Horvitz--Thompson estimators of the direct, indirect, and global average treatment effects under Bernoulli trials. We assume the heterogeneous additive treatment effect model with a hidden network that drives interference. Observing that these estimators are inconsistent in dense networks, we introduce novel eigenvector-based regression adjustment estimators to ensure consistency. We establish the asymptotic normality of the proposed estimators and provide conservative variance estimators under the design-based inference framework, offering robust conclusions independent of the underlying stochastic processes of the network and model parameters. Our method's adaptability is demonstrated across various interference structures, including partial interference and local interference in a two-sided marketplace. Numerical studies further illustrate the efficacy of the proposed estimators, offering practical insights into handling network interference.

Orbital magnetoelectric effect is closely related to the band topology of bulk crystalline insulators. Typical examples include the half quantized Chern-Simons orbital magnetoelectric coupling in three dimensional (3D) axion insulators and topological insulators, which are the hallmarks of their nontrivial bulk band topology. While the Chern-Simons coupling is well defined only for insulators with zero Chern number, the orbital magnetoelectric effects in 3D Chern insulators with nonzero (layer) Chern numbers are still open questions. In this work, we propose a never-mentioned quantization rule for the layer-resolved orbital magnetoelectric response in 3D Chern insulators, the gradient of which is exactly quantized in unit of $e^2/h$. By theoretical analysis and numerical simulations, we demonstrate that the quantized orbital magnetoelectric response remains robust for various types of interlayer hoppings and stackings, even against disorder and lack of symmetries. We argue that the robustness has a topological origin and protected by layer Chern number. It is thus promising to observe the proposed quantized orbital magnetoelectric response in a slab of 3D Chern insulator thanks to recent experimental developments., Comment: main text: 5 pages, 1 figure and 5 tables; SI: 15 pages, 5 figures and 2 tables

The vanadium-based kagome metals $A$V$_3$Sb$_5$ ($A$ = K, Rb, and Cs) host a superconducting ground state that coexists with an unconventional charge density wave (CDW). The CDW state exhibits experimental signatures of chirality, electronic nematicity, and time-reversal-symmetry-breaking, raising the questions whether the superconductivity (SC) in $A$V$_3$Sb$_5$ may also be unconventional, how SC interplays with CDW, and how the two orders evolve upon tuning. This article reviews studies of the superconducting pairing symmetry, and the tuning of SC and CDW in the $A$V$_3$Sb$_5$ compounds. Various experimental techniques consistently find that CsV$_3$Sb$_5$ exhibits nodeless SC, which remains robust regardless whether the CDW is present. Under hydrostatic pressure, SC in $A$V$_3$Sb$_5$ becomes enhanced as the CDW is gradually suppressed, revealing a competition between the two orders. In CsV$_3$Sb$_5$, a new CDW state emerges under pressure that competes more strongly with SC relative to the CDW at ambient pressure, and results in two superconducting domes that coexist with CDW. After the CDW in $A$V$_3$Sb$_5$ is fully suppressed with hydrostatic pressure, a further increase in pressure leads to a nonmonotonic evolution of the superconducting transition temperature driven by lattice modulations. Thickness is shown to be a powerful tuning parameter in $A$V$_3$Sb$_5$ thin flakes, revealing the evolution of CDW and SC upon dimensional reduction, and can be combined with hydrostatic pressure to shed light on the interplay between SC and CDW. Based on results reviewed in this article, we discuss outstanding issues to be addressed in the $A$V$_3$Sb$_5$ systems.

Online vectorized High-Definition (HD) map construction is crucial for subsequent prediction and planning tasks in autonomous driving. Following MapTR paradigm, recent works have made noteworthy achievements. However, reference points are randomly initialized in mainstream methods, leading to unstable matching between predictions and ground truth. To address this issue, we introduce PriorMapNet to enhance online vectorized HD map construction with priors. We propose the PPS-Decoder, which provides reference points with position and structure priors. Fitted from the map elements in the dataset, prior reference points lower the learning difficulty and achieve stable matching. Furthermore, we propose the PF-Encoder to enhance the image-to-BEV transformation with BEV feature priors. Besides, we propose the DMD cross-attention, which decouples cross-attention along multi-scale and multi-sample respectively to achieve efficiency. Our proposed PriorMapNet achieves state-of-the-art performance in the online vectorized HD map construction task on nuScenes and Argoverse2 datasets. The code will be released publicly soon.

Multi-turn dialogues are a key interaction method between humans and Large Language Models (LLMs), as conversations extend over multiple rounds, keeping LLMs' high generation quality and low latency is a challenge. Mainstream LLMs can be grouped into two categories based on masking strategy: causal LLM and prefix LLM. Several works have demonstrated that prefix LLMs tend to outperform causal ones in scenarios that heavily depend on historical context such as multi-turn dialogues or in-context learning, thanks to their bidirectional attention on prefix sequences. However, prefix LLMs have an inherent inefficient training problem in multi-turn dialogue datasets. In addition, the attention mechanism of prefix LLM makes it unable to reuse Key-Value Cache (KV Cache) across dialogue rounds to reduce generation latency. In this paper, we propose a novel masking scheme called Intermittent Semi-working Mask (ISM) to address these problems. Specifically, we apply alternate bidirectional and unidirectional attention on queries and answers in the dialogue history. In this way, ISM is able to maintain the high quality of prefix LLM and low generation latency of causal LLM, simultaneously. Extensive experiments illustrate that our ISM achieves significant performance.

In point cloud geometry compression, most octreebased context models use the cross-entropy between the onehot encoding of node occupancy and the probability distribution predicted by the context model as the loss. This approach converts the problem of predicting the number (a regression problem) and the position (a classification problem) of occupied child nodes into a 255-dimensional classification problem. As a result, it fails to accurately measure the difference between the one-hot encoding and the predicted probability distribution. We first analyze why the cross-entropy loss function fails to accurately measure the difference between the one-hot encoding and the predicted probability distribution. Then, we propose an attention-based child node number prediction (ACNP) module to enhance the context models. The proposed module can predict the number of occupied child nodes and map it into an 8- dimensional vector to assist the context model in predicting the probability distribution of the occupancy of the current node for efficient entropy coding. Experimental results demonstrate that the proposed module enhances the coding efficiency of octree-based context models., Comment: 2 figures and 2 tables

In point cloud geometry compression, context models usually use the one-hot encoding of node occupancy as the label, and the cross-entropy between the one-hot encoding and the probability distribution predicted by the context model as the loss function. However, this approach has two main weaknesses. First, the differences between contexts of different nodes are not significant, making it difficult for the context model to accurately predict the probability distribution of node occupancy. Second, as the one-hot encoding is not the actual probability distribution of node occupancy, the cross-entropy loss function is inaccurate. To address these problems, we propose a general structure that can enhance existing context models. We introduce the context feature residuals into the context model to amplify the differences between contexts. We also add a multi-layer perception branch, that uses the mean squared error between its output and node occupancy as a loss function to provide accurate gradients in backpropagation. We validate our method by showing that it can improve the performance of an octree-based model (OctAttention) and a voxel-based model (VoxelDNN) on the object point cloud datasets MPEG 8i and MVUB, as well as the LiDAR point cloud dataset SemanticKITTI., Comment: 11 pages, 8 figures

Learning-based methods have proven successful in compressing geometric information for point clouds. For attribute compression, however, they still lag behind non-learning-based methods such as the MPEG G-PCC standard. To bridge this gap, we propose a novel deep learning-based point cloud attribute compression method that uses a generative adversarial network (GAN) with sparse convolution layers. Our method also includes a module that adaptively selects the resolution of the voxels used to voxelize the input point cloud. Sparse vectors are used to represent the voxelized point cloud, and sparse convolutions process the sparse tensors, ensuring computational efficiency. To the best of our knowledge, this is the first application of GANs to compress point cloud attributes. Our experimental results show that our method outperforms existing learning-based techniques and rivals the latest G-PCC test model (TMC13v23) in terms of visual quality., Comment: 14 pages, 5 figures, Accepted by Computational Visual Media

The widespread applications of large language models (LLMs) have brought about concerns regarding their potential misuse. Although aligned with human preference data before release, LLMs remain vulnerable to various malicious attacks. In this paper, we adopt a red-teaming strategy to enhance LLM safety and introduce SoP, a simple yet effective framework to design jailbreak prompts automatically. Inspired by the social facilitation concept, SoP generates and optimizes multiple jailbreak characters to bypass the guardrails of the target LLM. Different from previous work which relies on proprietary LLMs or seed jailbreak templates crafted by human expertise, SoP can generate and optimize the jailbreak prompt in a cold-start scenario using open-sourced LLMs without any seed jailbreak templates. Experimental results show that SoP achieves attack success rates of 88% and 60% in bypassing the safety alignment of GPT-3.5-1106 and GPT-4, respectively. Furthermore, we extensively evaluate the transferability of the generated templates across different LLMs and held-out malicious requests, while also exploring defense strategies against the jailbreak attack designed by SoP. Code is available at https://github.com/Yang-Yan-Yang-Yan/SoP.

Efficient single instance segmentation is essential for unlocking features in the mobile imaging applications, such as capture or editing. Existing on-the-fly mobile imaging applications scope the segmentation task to portraits or the salient subject due to the computational constraints. Instance segmentation, despite its recent developments towards efficient networks, is still heavy due to the cost of computation on the entire image to identify all instances. To address this, we propose and formulate a one tap driven single instance segmentation task that segments a single instance selected by a user via a positive tap. This task, in contrast to the broader task of segmenting anything as suggested in the Segment Anything Model \cite{sam}, focuses on efficient segmentation of a single instance specified by the user. To solve this problem, we present TraceNet, which explicitly locates the selected instance by way of receptive field tracing. TraceNet identifies image regions that are related to the user tap and heavy computations are only performed on selected regions of the image. Therefore overall computation cost and memory consumption are reduced during inference. We evaluate the performance of TraceNet on instance IoU average over taps and the proportion of the region that a user tap can fall into for a high-quality single-instance mask. Experimental results on MS-COCO and LVIS demonstrate the effectiveness and efficiency of the proposed approach. TraceNet can jointly achieve the efficiency and interactivity, filling in the gap between needs for efficient mobile inference and recent research trend towards multimodal and interactive segmentation models.

The personalized bundle generation problem, which aims to create a preferred bundle for user from numerous candidate items, receives increasing attention in recommendation. However, existing works ignore the order-invariant nature of the bundle and adopt sequential modeling methods as the solution, which might introduce inductive bias and cause a large latency in prediction. To address this problem, we propose to perform the bundle generation via non-autoregressive mechanism and design a novel encoder-decoder framework named BundleNAT, which can effectively output the targeted bundle in one-shot without relying on any inherent order. In detail, instead of learning sequential dependency, we propose to adopt pre-training techniques and graph neural network to fully embed user-based preference and item-based compatibility information, and use a self-attention based encoder to further extract global dependency pattern. We then design a permutation-equivariant decoding architecture that is able to directly output the desired bundle in a one-shot manner. Experiments on three real-world datasets from Youshu and Netease show the proposed BundleNAT significantly outperforms the current state-of-the-art methods in average by up to 35.92%, 10.97% and 23.67% absolute improvements in Precision, Precision+, and Recall, respectively., Comment: Submitted to Information Processing & Management

Fractional quantum anomalous hall effect (FQAHE), a transport effect with fractionally quantized Hall plateau emerging under zero magnetic field, provides a radically new opportunity to engineer topological quantum electronics. By construction of topological flat band with moire engineering, intrinsic FQAHE has been observed in twisted MoTe2 system and rhombohedral pentalayer graphene/hBN moire superlattices with anomalous Hall resistivity quantization number C <= 2/3 including the gapless composite Fermi-liquid state with C = 1/2. Here we experimentally demonstrate a new system of rhombohedral hexalayer graphene (RHG)/hBN moire superlattices showing both fractional and integer quantum anomalous Hall effects when the lowest flat Chern band is fractionally and fully filled at zero magnetic field. The zero-field Hall resistance Rho_xy = h/Ce2 is quantized to values corresponding to C = 3/5, 2/3, 5/7, 3/4, 7/9 and 1 at moire filling factors v = 3/5, 2/3, 5/7, 3/4, 7/9 and 1, respectively. Particularly, the C = 3/4 FQAHE state at v = 3/4 moire filling featuring a minimum of longitudinal resistance Rho_xx and fractionally quantized Hall resistance Rho_xy = 4h/3e2, is observed for the first time under zero magnetic field. Such a state may be similar to the C = 3/4 fractional quantum hall (FQHE) state recently observed at high magnetic fields9,10 and possibly host fractional charge excitations obeying non-Abelian statistics. By tuning the electrical and magnetic fields at 0 < v < 1, we have observed a sign reversal of the Hall resistivity for v = 2/3 state, indicating a transition from quasi-electron-like excitations to quasi-hole ones. Our experiment has established RHG/hBN moire superlattices a promising platform to explore quasi-particles with fractional charge excitations and non-Abelian anyons at zero magnetic field.

Let $S$ be a minimal irregular surface of general type, whose Albanese map induces a fibration $f:\,S \to C$ of genus $g$.We prove a linear upper bound on the genus $g$ if $K_S^2\leq 4\chi(\mathcal{O}_S)$. Examples are constructed showing that the above linear upper bound is sharp. We also give a characterization of the Albanese fibrations reaching the above upper bound when $\chi(\mathcal{O}_S)\geq 5$.On the other hand, we will construct a sequence of surfaces $S_n$ of general type with $K_{S_n}^2/\chi(\mathcal{O}_{S_n})>4$ and with an Albanese fibration $f_n$, such that the genus $g_n$ of a general fiber of $f_n$ increases quadratically with $\chi(\mathcal{O}_{S_n})$,and that $K_{S_n}^2/\chi(\mathcal{O}_{S_n})$ can be arbitrarily close to $4$., Comment: Add a characterization of the Albanese fibrations reaching the above upper bound. Comments are welcome!

In the rapidly evolving automotive industry, Systems-on-Chips (SoCs) are playing an increasingly crucial role in enhancing vehicle intelligence, connectivity, and safety features. For enterprises whose business encompasses automotive SoCs, the sustained and stable provision and receipt of SoC relevant goods or services are essential. Considering the imperative for a resilient and adaptable supply network, enterprises are concentrating their efforts on formulating strategies to address risks stemming from supply chain disruptions caused by technological obsolescence, natural disasters, and geopolitical tensions. This study presents an open supply knowledge extraction and complement approach and build a supply chain network of automotive SoC enterprises in China, which incorporates cross-domain named entity recognition under limited information, fuzzy matching of firm entities, and supply relation inferring based on knowledge graph. Subsequently, we exhibit the degree and registered capital distribution across firms, and analyze the correlations between centrality metrics in the supply chain network. Finally, based on recovery capacity and risk transfer, two interaction disruption models (IDMs) are developed to elucidate the adaptive behaviors and effect of network disruptions under various business and attack strategies. This research not only aids in exploring the complexities of Chinese automotive SoC supply chain but also enriches our understanding of the dynamics of firm behavior in this crucial industry sector., Comment: arXiv admin note: text overlap with arXiv:2304.10428 by other authors

Social media aids disaster response but suffers from noise, hindering accurate impact assessment and decision making for resilient cities, which few studies considered. To address the problem, this study proposes the first domain-specific LLM model and an integrated method for rapid earthquake impact assessment. First, a few categories are introduced to classify and filter microblogs considering their relationship to the physical and social impacts of earthquakes, and a dataset comprising 7282 earthquake-related microblogs from twenty earthquakes in different locations is developed as well. Then, with a systematic analysis of various influential factors, QuakeBERT, a domain-specific large language model (LLM), is developed and fine-tuned for accurate classification and filtering of microblogs. Meanwhile, an integrated method integrating public opinion trend analysis, sentiment analysis, and keyword-based physical impact quantification is introduced to assess both the physical and social impacts of earthquakes based on social media texts. Experiments show that data diversity and data volume dominate the performance of QuakeBERT and increase the macro average F1 score by 27%, while the best classification model QuakeBERT outperforms the CNN- or RNN-based models by improving the macro average F1 score from 60.87% to 84.33%. Finally, the proposed approach is applied to assess two earthquakes with the same magnitude and focal depth. Results show that the proposed approach can effectively enhance the impact assessment process by accurate detection of noisy microblogs, which enables effective post-disaster emergency responses to create more resilient cities.

Kodaira fibrations have non-trivial vertical fundamental groups and their slopes are all $12$. In this paper, we show that $12$ is indeed the sharp upper bound for the slopes of fibrations with trivial vertical fundamental groups. Precisely, for each $g\geq3$ we prove the existence of fibrations of genus $g$ with trivial vertical fundamental groups whose slopes can be arbitrarily close to $12$. This gives a relative analogy of Roulleau-Urz\'ua's work on the slopes of surfaces of general type with trivial fundamental groups., Comment: Any comment is warmly welcome

How intelligence emerges from living beings has been a fundamental question in neuroscience. However, it remains largely unanswered due to the complex neuronal dynamics and intricate connections between neurons in real neural systems. To address this challenge, we leveraged the largest available adult Drosophila connectome data set, and constructed a comprehensive computational framework based on simplified neuronal activation mechanisms to simulate the observed activation behavior within the connectome. The results revealed that even with rudimentary neuronal activation mechanisms, models grounded in real neural network structures can generate activation patterns strikingly similar to those observed in the actual brain. A significant discovery was the consistency of activation patterns across various neuronal dynamic models. This consistency, achieved with the same network structure, underscores the pivotal role of network topology in neural information processing. These results challenge the prevailing view that solely relies on neuron count or complex individual neuron dynamics. Further analysis demonstrated a near-complete separation of the visual and olfactory systems at the network level. Moreover, we found that the network distance, rather than spatial distance, is the primary determinant of activation patterns. Additionally, our experiments revealed that a reconnect rate of at least 0.1% was sufficient to disrupt the previously observed activation patterns. We also observed synergistic effects between the brain hemispheres: Even with unilateral input stimuli, visual-related neurons in both hemispheres were activated, highlighting the importance of interhemispheric communication. These findings emphasize the crucial role of network structure in neural activation and offer novel insights into the fundamental principles governing brain functionality.

Ring artifacts in computed tomography images, arising from the undesirable responses of detector units, significantly degrade image quality and diagnostic reliability. To address this challenge, we propose a dual-domain regularization model to effectively remove ring artifacts, while maintaining the integrity of the original CT image. The proposed model corrects the vertical stripe artifacts on the sinogram by innovatively updating the response inconsistency compensation coefficients of detector units, which is achieved by employing the group sparse constraint and the projection-view direction sparse constraint on the stripe artifacts. Simultaneously, we apply the sparse constraint on the reconstructed image to further rectified ring artifacts in the image domain. The key advantage of the proposed method lies in considering the relationship between the response inconsistency compensation coefficients of the detector units and the projection views, which enables a more accurate correction of the response of the detector units. An alternating minimization method is designed to solve the model. Comparative experiments on real photon counting detector data demonstrate that the proposed method not only surpasses existing methods in removing ring artifacts but also excels in preserving structural details and image fidelity.

We show that there does not exist any Shimura curve with strictly maximal Higgs field generically in the Torelli locus of non-hyperelliptic curves of genus $g\geq 4$. In particular, Shimura curves of Mumford type are not generically in the Torelli locus of non-hyperelliptic curves of genus $g\geq 4$., Comment: Any comment is welcome

Pre-trained language models have been proven to possess strong base capabilities, which not only excel in in-distribution language modeling but also show powerful abilities in out-of-distribution language modeling, transfer learning and few-shot learning. Unlike existing work focusing on the influence of scale on base capabilities, our work examines the influence of architecture on those. Specifically, our concern is: How does architecture influence the base capabilities of pre-trained language models? In this work, we attempt to explain and reverse the decline in base capabilities caused by the architecture of FFN-Wider Transformers, seeking to provide some insights. Through analysis, we found the contribution ratio of Multi-Head Attention (a combination function) to pre-trained language modeling is a key factor affecting base capabilities. FFN-Wider Transformers reduce the contribution ratio of this combination function, leading to a decline in base capabilities. We confirmed this by experiments and proposed Combination Enhanced Architecture (CEA) to address the decline in base capabilities of such models. Significantly, we extended our explanation and CEA to Mixture of Experts (MoE) Transformers. We successfully achieved significant improvements in base capabilities on a 14B parameter MoE model, demonstrating the practical application value of our work. This also indicates that our analysis has a certain guiding significance for architecture analysis, architecture improvement and architecture design., Comment: Accepted by NeurIPS 2024

Recently, Mixture of Experts (MoE) Transformers have garnered increasing attention due to their advantages in model capacity and computational efficiency. However, studies have indicated that MoE Transformers underperform vanilla Transformers in many downstream tasks, significantly diminishing the practical value of MoE models. To explain this issue, we propose that the pre-training performance and transfer capability of a model are joint determinants of its downstream task performance. MoE models, in comparison to vanilla models, have poorer transfer capability, leading to their subpar performance in downstream tasks. To address this issue, we introduce the concept of transfer capability distillation, positing that although vanilla models have weaker performance, they are effective teachers of transfer capability. The MoE models guided by vanilla models can achieve both strong pre-training performance and transfer capability, ultimately enhancing their performance in downstream tasks. We design a specific distillation method and conduct experiments on the BERT architecture. Experimental results show a significant improvement in downstream performance of MoE models, and many further evidences also strongly support the concept of transfer capability distillation. Finally, we attempt to interpret transfer capability distillation and provide some insights from the perspective of model feature.