Search

Recovery of a sparse signal from a nonlinear system arises in many practical applications including compressive sensing, image reconstruction and machine learning. In this paper, a fast block nonlinear Bregman-Kaczmarz method with averaging is presented for nonlinear sparse signal recovery problems. Theoretical analysis proves that the averaging block nonlinear Bregman-Kaczmarz method with both constant stepsizes and adaptive stepsizes are convergent. Numerical experiments demonstrate the effectiveness of the averaging block nonlinear Bregman-Kaczmarz method, which converges faster than the existing nonlinear Bregman-Kaczmarz methods.

We consider a natural class of extensions of the Anderson model on $\mathbb Z^d$, called random block Schr\"odinger operators (RBSOs), which feature a random block potential. They are defined on the $d$-dimensional lattice $(\mathbb Z/L\mathbb Z)^d$ and take the form $H=V + \lambda \Psi$, where $V$ is a diagonal block matrix with diagonal blocks that are i.i.d. $W^d\times W^d$ GUE, $\Psi$ describes the interactions between neighboring blocks, and $0<\lambda\ll 1$ is a small coupling parameter (making $H$ a perturbation of $V$). We focus on three specific RBSOs: (1) the block Anderson model, where $\Psi$ is the Laplacian operator on $\mathbb Z^d$; (2) the Anderson orbital model, where $\Psi$ is a block Laplacian operator; (3) the Wegner orbital model, where the nearest-neighbor blocks of $\Psi$ are random matrices. Assuming $d\ge 7$ and $W\ge L^\varepsilon$ for a small constant $\varepsilon>0$, under certain lower bound on $\lambda$, we prove delocalization and quantum unique ergodicity for the bulk eigenvectors, as well as a quantum diffusion estimate for the Green's function. Together with the localization result established in arXiv:1608.02922, our results rigorously demonstrate the existence of an Anderson localization-delocalization transition for RBSOs as $\lambda$ varies. Our proof is based on the $T$-expansion method and the concept of self-energy renormalization developed for random band matrices in arXiv:2104.12048. Additionally, we explore a conceptually new idea called coupling renormalization. This phenomenon extends the self-energy renormalization and is prevalent in quantum field theory, yet it is identified here for the first time within the context of random Schr\"odinger operators. We anticipate that our techniques can be adapted to real or non-Gaussian block potentials and more general interactions $\Psi$., Comment: 123 pages. Minor updates

Recent advancements in text-to-image (T2I) generation using diffusion models have enabled cost-effective video-editing applications by leveraging pre-trained models, eliminating the need for resource-intensive training. However, the frame-independence of T2I generation often results in poor temporal consistency. Existing methods address this issue through temporal layer fine-tuning or inference-based temporal propagation, but these approaches suffer from high training costs or limited temporal coherence. To address these challenges, we propose a General and Efficient Adapter (GE-Adapter) that integrates temporal-spatial and semantic consistency with Baliteral DDIM inversion. This framework introduces three key components: (1) Frame-based Temporal Consistency Blocks (FTC Blocks) to capture frame-specific features and enforce smooth inter-frame transitions via temporally-aware loss functions; (2) Channel-dependent Spatial Consistency Blocks (SCD Blocks) employing bilateral filters to enhance spatial coherence by reducing noise and artifacts; and (3) Token-based Semantic Consistency Module (TSC Module) to maintain semantic alignment using shared prompt tokens and frame-specific tokens. Our method significantly improves perceptual quality, text-image alignment, and temporal coherence, as demonstrated on the MSR-VTT dataset. Additionally, it achieves enhanced fidelity and frame-to-frame coherence, offering a practical solution for T2V editing.

Consider an $ N \times N$ Hermitian one-dimensional random band matrix with band width $W > N^{1 / 2 + \varepsilon} $ for any $ \varepsilon > 0$. In the bulk of the spectrum and in the large $ N $ limit, we obtain the following results: (i) The semicircle law holds up to the scale $ N^{-1 + \varepsilon} $ for any $ \varepsilon > 0 $. (ii) All $ L^2 $- normalized eigenvectors are delocalized, meaning their $ L^\infty$ norms are simultaneously bounded by $ N^{-\frac{1}{2} + \varepsilon} $ with overwhelming probability, for any $ \varepsilon > 0 $. (iii) Quantum unique ergodicity holds in the sense that the local $ L^2 $ mass of eigenvectors becomes equidistributed with high probability. (iv) Universality of eigenvalue statistics, i.e., the local eigenvalue statistics of these band matrices are given by those of Gaussian unitary ensembles., Comment: 86 pages, 14 figures, minor revision

Understanding the complex interactions between land surface and atmosphere is essential to improve weather and climate predictions. Various numerical experiments have suggested that regions of strong coupling strength (hotspots) are located in the transitional climate zones. However, atmospheric processes in these hotspots are found to have different responses to the perturbation of surface properties. Here we establish analytical relationships to identify key role of soil moisture variances in controlling the coupling hotspots. Using the most recent numerical experiments, we find different signs of feedback in two such hotspots, suggesting the coupling can either reinforce or attenuate persistent extreme climates. We further uncover new coupling hotspots in regions where precipitation is highly sensitive to soil moisture perturbation. Our results highlight the importance of both signs and magnitudes of land-atmosphere interactions over extensive regions, where the ecosystems and communities are particularly vulnerable to the extreme climate events., Comment: 3 figures, 18 pages

Multimodal attributed graphs (MAGs) are prevalent in various real-world scenarios and generally contain two kinds of knowledge: (a) Attribute knowledge is mainly supported by the attributes of different modalities contained in nodes (entities) themselves, such as texts and images. (b) Topology knowledge, on the other hand, is provided by the complex interactions posed between nodes. The cornerstone of MAG representation learning lies in the seamless integration of multimodal attributes and topology. Recent advancements in Pre-trained Language/Vision models (PLMs/PVMs) and Graph neural networks (GNNs) have facilitated effective learning on MAGs, garnering increased research interest. However, the absence of meaningful benchmark datasets and standardized evaluation procedures for MAG representation learning has impeded progress in this field. In this paper, we propose Multimodal Attribute Graph Benchmark (MAGB)}, a comprehensive and diverse collection of challenging benchmark datasets for MAGs. The MAGB datasets are notably large in scale and encompass a wide range of domains, spanning from e-commerce networks to social networks. In addition to the brand-new datasets, we conduct extensive benchmark experiments over MAGB with various learning paradigms, ranging from GNN-based and PLM-based methods, to explore the necessity and feasibility of integrating multimodal attributes and graph topology. In a nutshell, we provide an overview of the MAG datasets, standardized evaluation procedures, and present baseline experiments. The entire MAGB project is publicly accessible at https://github.com/sktsherlock/ATG.

Owing to the unprecedented capability in semantic understanding and logical reasoning, the pre-trained large language models (LLMs) have shown fantastic potential in developing the next-generation recommender systems (RSs). However, the static index paradigm adopted by current methods greatly restricts the utilization of LLMs capacity for recommendation, leading to not only the insufficient alignment between semantic and collaborative knowledge, but also the neglect of high-order user-item interaction patterns. In this paper, we propose Twin-Tower Dynamic Semantic Recommender (TTDS), the first generative RS which adopts dynamic semantic index paradigm, targeting at resolving the above problems simultaneously. To be more specific, we for the first time contrive a dynamic knowledge fusion framework which integrates a twin-tower semantic token generator into the LLM-based recommender, hierarchically allocating meaningful semantic index for items and users, and accordingly predicting the semantic index of target item. Furthermore, a dual-modality variational auto-encoder is proposed to facilitate multi-grained alignment between semantic and collaborative knowledge. Eventually, a series of novel tuning tasks specially customized for capturing high-order user-item interaction patterns are proposed to take advantages of user historical behavior. Extensive experiments across three public datasets demonstrate the superiority of the proposed methodology in developing LLM-based generative RSs. The proposed TTDS recommender achieves an average improvement of 19.41% in Hit-Rate and 20.84% in NDCG metric, compared with the leading baseline methods.

Learning from interpretation transition (LFIT) is a framework for learning rules from observed state transitions. LFIT has been implemented in purely symbolic algorithms, but they are unable to deal with noise or generalize to unobserved transitions. Rule extraction based neural network methods suffer from overfitting, while more general implementation that categorize rules suffer from combinatorial explosion. In this paper, we introduce a technique to leverage variable permutation invariance inherent in symbolic domains. Our technique ensures that the permutation and the naming of the variables would not affect the results. We demonstrate the effectiveness and the scalability of this method with various experiments. Our code is publicly available at https://github.com/phuayj/delta-lfit-2

This paper introduces Lite-SAM, an efficient end-to-end solution for the SegEvery task designed to reduce computational costs and redundancy. Lite-SAM is composed of four main components: a streamlined CNN-Transformer hybrid encoder (LiteViT), an automated prompt proposal network (AutoPPN), a traditional prompt encoder, and a mask decoder. All these components are integrated within the SAM framework. Our LiteViT, a high-performance lightweight backbone network, has only 1.16M parameters, which is a 23% reduction compared to the lightest existing backbone network Shufflenet. We also introduce AutoPPN, an innovative end-to-end method for prompt boxes and points generation. This is an improvement over traditional grid search sampling methods, and its unique design allows for easy integration into any SAM series algorithm, extending its usability. we have thoroughly benchmarked Lite-SAM across a plethora of both public and private datasets. The evaluation encompassed a broad spectrum of universal metrics, including the number of parameters, SegEvery execution time, and accuracy. The findings reveal that Lite-SAM, operating with a lean 4.2M parameters, significantly outpaces its counterparts, demonstrating performance improvements of 43x, 31x, 20x, 21x, and 1.6x over SAM, MobileSAM, Edge-SAM, EfficientViT-SAM, and MobileSAM-v2 respectively, all the while maintaining competitive accuracy. This underscores Lite-SAM's prowess in achieving an optimal equilibrium between performance and precision, thereby setting a new state-of-the-art(SOTA) benchmark in the domain., Comment: ECCV 2024 Accepted

The Sparse Kaczmarz method is a famous and widely used iterative method for solving the regularized basis pursuit problem. A general scheme of the surrogate hyperplane sparse Kaczmarz method is proposed. In particular, a class of residual-based surrogate hyperplane sparse Kaczmarz method is introduced and the implementations are well discussed. Their convergence theories are proved and the linear convergence rates are studied and compared in details. Numerical experiments verify the efficiency of the proposed methods.

Spatio-Temporal Convolutional Neural Networks (ST-CNN) allow extending CNN capabilities from image processing to consecutive temporal-pattern recognition. Generally, state-of-the-art (SotA) ST-CNNs inflate the feature maps and weights from well-known CNN backbones to represent the additional time dimension. However, edge computing applications would suffer tremendously from such large computation or memory overhead. Fortunately, the overlapping nature of ST-CNN enables various optimizations, such as the dilated causal convolution structure and Depth-First (DF) layer fusion to reuse the computation between time steps and CNN sliding windows, respectively. Yet, no hardware-aware approach has been proposed that jointly explores the optimal strategy from a scheduling as well as a hardware point of view. To this end, we present ACCO, an automated optimizer that explores efficient Causal CNN transformation and DF scheduling for ST-CNNs on edge hardware accelerators. By cost-modeling the computation and data movement on the accelerator architecture, ACCO automatically selects the best scheduling strategy for the given hardware-algorithm target. Compared to the fixed dilated causal structure, ST-CNNs with ACCO reach an ~8.4x better Energy-Delay-Product. Meanwhile, ACCO improves ~20% in layer-fusion optimals compared to the SotA DF exploration toolchain. When jointly optimizing ST-CNN on the temporal and spatial dimension, ACCO's scheduling outcomes are on average 19x faster and 37x more energy-efficient than spatial DF schemes.

The land-atmosphere coupling strength has been defined as the percentage of precipitation variability explained by the variation of soil moisture in the Global Land-Atmosphere Coupling Experiment (GLACE). While it is useful to identify global hotspots of land-atmosphere interaction, this coupling strength is different from coupling sensitivity, which directly quantifies how precipitation generation responds to the perturbation of soil moisture and is essential for our understanding of the global water cycle. To disentangle these two quantities, here we theoretically explore the relationships among coupling strength, sensitivity, and soil moisture variances. We use climate model outputs to show that the largest soil moisture variances are located in the transitional climate zones and the variations of soil moisture largely account for the geographical patterns of coupling hotspots. The coupling sensitivity is not necessarily low in non-hotspot regions, which could impose great impacts on the development of extreme climate events. We therefore call for more research attention on coupling sensitivity to improve our understanding of the climate system.

Graph Neural Networks (GNNs) have achieved notable success in various applications over graph data. However, recent research has revealed that real-world graphs often contain noise, and GNNs are susceptible to noise in the graph. To address this issue, several Graph Structure Learning (GSL) models have been introduced. While GSL models are tailored to enhance robustness against edge noise through edge reconstruction, a significant limitation surfaces: their high reliance on node features. This inherent dependence amplifies their susceptibility to noise within node features. Recognizing this vulnerability, we present DEGNN, a novel GNN model designed to adeptly mitigate noise in both edges and node features. The core idea of DEGNN is to design two separate experts: an edge expert and a node feature expert. These experts utilize self-supervised learning techniques to produce modified edges and node features. Leveraging these modified representations, DEGNN subsequently addresses downstream tasks, ensuring robustness against noise present in both edges and node features of real-world graphs. Notably, the modification process can be trained end-to-end, empowering DEGNN to adjust dynamically and achieves optimal edge and node representations for specific tasks. Comprehensive experiments demonstrate DEGNN's efficacy in managing noise, both in original real-world graphs and in graphs with synthetic noise., Comment: PAKDD 2024, the code is available at https://github.com/TaiHasegawa/DEGNN

Exemplar-Free Class Incremental Learning is a highly challenging setting where replay memory is unavailable. Methods relying on frozen feature extractors have drawn attention recently in this setting due to their impressive performances and lower computational costs. However, those methods are highly dependent on the data used to train the feature extractor and may struggle when an insufficient amount of classes are available during the first incremental step. To overcome this limitation, we propose to use a pre-trained text-to-image diffusion model in order to generate synthetic images of future classes and use them to train the feature extractor. Experiments on the standard benchmarks CIFAR100 and ImageNet-Subset demonstrate that our proposed method can be used to improve state-of-the-art methods for exemplar-free class incremental learning, especially in the most difficult settings where the first incremental step only contains few classes. Moreover, we show that using synthetic samples of future classes achieves higher performance than using real data from different classes, paving the way for better and less costly pre-training methods for incremental learning.

We consider a constant-size subset of left and right eigenvectors of an $N\times N$ i.i.d. complex non-Hermitian matrix associated with the eigenvalues with pairwise distances at least $N^{-\frac12+\epsilon}$. We show that arbitrary constant rank projections of these eigenvectors are Gaussian and jointly independent., Comment: 46 pages

In the pursuit of novel catalyst development to address pressing environmental concerns and energy demand, conventional design and optimization methods often fall short due to the complexity and vastness of the catalyst parameter space. The advent of Machine Learning (ML) has ushered in a new era in the field of catalyst optimization, offering potential solutions to the shortcomings of traditional techniques. However, existing methods fail to effectively harness the wealth of information contained within the burgeoning body of scientific literature on catalyst synthesis. To address this gap, this study proposes an innovative Artificial Intelligence (AI) workflow that integrates Large Language Models (LLMs), Bayesian optimization, and an active learning loop to expedite and enhance catalyst optimization. Our methodology combines advanced language understanding with robust optimization strategies, effectively translating knowledge extracted from diverse literature into actionable parameters for practical experimentation and optimization. In this article, we demonstrate the application of this AI workflow in the optimization of catalyst synthesis for ammonia production. The results underscore the workflow's ability to streamline the catalyst development process, offering a swift, resource-efficient, and high-precision alternative to conventional methods., Comment: 31 pages, 7 figures

The leading operational Global Ocean Forecasting Systems (GOFSs) use physics-driven numerical forecasting models that solve the partial differential equations with expensive computation. Recently, specifically in atmosphere weather forecasting, data-driven models have demonstrated significant potential for speeding up environmental forecasting by orders of magnitude, but there is still no data-driven GOFS that matches the forecasting accuracy of the numerical GOFSs. In this paper, we propose the first data-driven 1/12{\deg} resolution global ocean eddy-resolving forecasting model named XiHe, which is established from the 25-year France Mercator Ocean International's daily GLORYS12 reanalysis data. XiHe is a hierarchical transformer-based framework coupled with two special designs. One is the land-ocean mask mechanism for focusing exclusively on the global ocean circulation. The other is the ocean-specific block for effectively capturing both local ocean information and global teleconnection. Extensive experiments are conducted under satellite observations, in situ observations, and the IV-TT Class 4 evaluation framework of the world's leading operational GOFSs from January 2019 to December 2020. The results demonstrate that XiHe achieves stronger forecast performance in all testing variables than existing leading operational numerical GOFSs including Mercator Ocean Physical SYstem (PSY4), Global Ice Ocean Prediction System (GIOPS), BLUElinK OceanMAPS (BLK), and Forecast Ocean Assimilation Model (FOAM). Particularly, the accuracy of ocean current forecasting of XiHe out to 60 days is even better than that of PSY4 in just 10 days. Additionally, XiHe is able to forecast the large-scale circulation and the mesoscale eddies. Furthermore, it can make a 10-day forecast in only 0.35 seconds, which accelerates the forecast speed by thousands of times compared to the traditional numerical GOFSs.

Dynamic control of circular dichroism in photonic structures is critically important for compact spectrometers, stereoscopic displays, and information processing exploiting multiple degrees of freedom. Metasurfaces can help miniaturize chiral devices but only produce static and limited chiral responses. While external stimuli are able to tune resonances, their modulations are often weak, and reversing continuously the sign of circular dichroism is extremely challenging. Here, we demonstrate dynamically tunable chiral response of resonant metasurfaces supporting chiral bound states in the continuum combining them with phase-change materials. Phase transition between amorphous and crystalline phases allows to control chiral response and vary chirality rapidly from -0.947 to +0.958 backward and forward via chirality continuum. Our demonstrations underpin the rapid development of chiral photonics and its applications., Comment: 14 pages, 4 figures

Consider $D$ random systems that are modeled by independent $N\times N$ complex Hermitian Wigner matrices. Suppose they are lying on a circle and the neighboring systems interact with each other through a deterministic matrix $A$. We prove that in the asymptotic limit $N\to \infty$, the whole system exhibits a quantum chaos transition when the interaction strength $\|A\|_{HS}$ varies. Specifically, when $\|A\|_{HS}\ge N^{\varepsilon}$, we prove that the bulk eigenvalue statistics match those of a $DN\times DN$ GUE asymptotically and each bulk eigenvector is approximately equally distributed among the $D$ subsystems with probability $1-o(1)$. These phenomena indicate quantum chaos of the whole system. In contrast, when $\|A\|_{HS}\le N^{-\varepsilon}$, we show that the system is integrable: the bulk eigenvalue statistics behave like $D$ independent copies of GUE statistics asymptotically and each bulk eigenvector is localized on only one subsystem. In particular, if we take $D\to \infty$ after the $N\to \infty$ limit, the bulk statistics converge to a Poisson point process under the $DN$ scaling., Comment: Final version. Accepted by Communications in Mathematical Physics

Using ultrafast broad-band transient absorption (TA) spectroscopy of photo-excited MAPbBr3 thin films with probe continua in the visible and the mid-to-deep-UV ranges, we capture the ultrafast gap renormalization at the fundamental gap situated at the R symmetry point of the Brillouin Zone (BZ) and a higher energy gap at the M point. Global Lifetime analysis, Lifetime density distribution and spectral trace fitting analysis are applied to extract quantitative information. Our work confirms the similarity of the energy gap renormalization at both symmetry points, which rises within the instrument response function (IRF, ~250 fs) and decays in ~400-600 fs times, and an energy red-shift of ~90-150 meV. The ability to monitor different high symmetry points in photoexcited perovskites opens exciting prospects into the characterization of materials, which can be extended to a larger class of materials., Comment: 16 pages, 4 figures