Search

Semi-supervised change detection (SSCD) utilizes partially labeled data and a large amount of unlabeled data to detect changes. However, the transformer-based SSCD network does not perform as well as the convolution-based SSCD network due to the lack of labeled data. To overcome this limitation, we introduce a new decoder called Cross Branch Feature Fusion CBFF, which combines the strengths of both local convolutional branch and global transformer branch. The convolutional branch is easy to learn and can produce high-quality features with a small amount of labeled data. The transformer branch, on the other hand, can extract global context features but is hard to learn without a lot of labeled data. Using CBFF, we build our SSCD model based on a strong-to-weak consistency strategy. Through comprehensive experiments on WHU-CD and LEVIR-CD datasets, we have demonstrated the superiority of our method over seven state-of-the-art SSCD methods., Comment: 5 pages, 4 figures, accepted by ICASSP 2024

Aspect-Based Sentiment Analysis (ABSA) in tourism plays a significant role in understanding tourists' evaluations of specific aspects of attractions, which is crucial for driving innovation and development in the tourism industry. However, traditional pipeline models are afflicted by issues such as error propagation and incomplete extraction of sentiment elements. To alleviate this issue, this paper proposes an aspect-based sentiment analysis model, ACOS_LLM, for Aspect-Category-Opinion-Sentiment Quadruple Extraction (ACOSQE). The model comprises two key stages: auxiliary knowledge generation and ACOSQE. Firstly, Adalora is used to fine-tune large language models for generating high-quality auxiliary knowledge. To enhance model efficiency, Sparsegpt is utilized to compress the fine-tuned model to 50% sparsity. Subsequently, Positional information and sequence modeling are employed to achieve the ACOSQE task, with auxiliary knowledge and the original text as inputs. Experiments are conducted on both self-created tourism datasets and publicly available datasets, Rest15 and Rest16. Results demonstrate the model's superior performance, with an F1 improvement of 7.49% compared to other models on the tourism dataset. Additionally, there is an F1 improvement of 0.05% and 1.06% on the Rest15 and Rest16 datasets, respectively., Comment: 19 pages, 17 figures

Large Language Models (LLMs) have excelled in various tasks but are still vulnerable to jailbreaking attacks, where attackers create jailbreak prompts to mislead the model to produce harmful or offensive content. Current jailbreak methods either rely heavily on manually crafted templates, which pose challenges in scalability and adaptability, or struggle to generate semantically coherent prompts, making them easy to detect. Additionally, most existing approaches involve lengthy prompts, leading to higher query costs.In this paper, to remedy these challenges, we introduce a novel jailbreaking attack framework, which is an automated, black-box jailbreaking attack framework that adapts the black-box fuzz testing approach with a series of customized designs. Instead of relying on manually crafted templates, our method starts with an empty seed pool, removing the need to search for any related jailbreaking templates. We also develop three novel question-dependent mutation strategies using an LLM helper to generate prompts that maintain semantic coherence while significantly reducing their length. Additionally, we implement a two-level judge module to accurately detect genuine successful jailbreaks. We evaluated our method on 7 representative LLMs and compared it with 5 state-of-the-art jailbreaking attack strategies. For proprietary LLM APIs, such as GPT-3.5 turbo, GPT-4, and Gemini-Pro, our method achieves attack success rates of over 90%, 80%, and 74%, respectively, exceeding existing baselines by more than 60%. Additionally, our method can maintain high semantic coherence while significantly reducing the length of jailbreak prompts. When targeting GPT-4, our method can achieve over 78\% attack success rate even with 100 tokens. Moreover, our method demonstrates transferability and is robust to state-of-the-art defenses. We will open-source our codes upon publication.

Finger vein recognition technology has become one of the primary solutions for high-security identification systems. However, it still has information leakage problems, which seriously jeopardizes users privacy and anonymity and cause great security risks. In addition, there is no work to consider a fully integrated secure finger vein recognition system. So, different from the previous systems, we integrate preprocessing and template protection into an integrated deep learning model. We propose an end-to-end cancelable finger vein network (CFVNet), which can be used to design an secure finger vein recognition system.It includes a plug-and-play BWR-ROIAlign unit, which consists of three sub-modules: Localization, Compression and Transformation. The localization module achieves automated localization of stable and unique finger vein ROI. The compression module losslessly removes spatial and channel redundancies. The transformation module uses the proposed BWR method to introduce unlinkability, irreversibility and revocability to the system. BWR-ROIAlign can directly plug into the model to introduce the above features for DCNN-based finger vein recognition systems. We perform extensive experiments on four public datasets to study the performance and cancelable biometric attributes of the CFVNet-based recognition system. The average accuracy, EERs and Dsys on the four datasets are 99.82%, 0.01% and 0.025, respectively, and achieves competitive performance compared with the state-of-the-arts.

The evaluation of the problem-solving capability under incomplete information scenarios of Large Language Models (LLMs) is increasingly important, encompassing capabilities such as questioning, knowledge search, error detection, and path planning. Current research mainly focus on LLMs' problem-solving capability such as ``Twenty Questions''. However, these kinds of games do not require recognizing misleading cues which are necessary in the incomplete information scenario. Moreover, the existing game such as ``Who is undercover'' are highly subjective, making it challenging for evaluation. Therefore, in this paper, we introduce a novel game named BrainKing based on the ``Who is undercover'' and ``Twenty Questions'' for evaluating LLM capabilities under incomplete information scenarios. It requires LLMs to identify target entities with limited yes-or-no questions and potential misleading answers. By setting up easy, medium, and hard difficulty modes, we comprehensively assess the performance of LLMs across various aspects. Our results reveal the capabilities and limitations of LLMs in BrainKing, providing significant insights of LLM problem-solving levels., Comment: Accepted to ACL 2024 (Findings)

The sixth-generation (6G) non-terrestrial networks (NTNs) are crucial for real-time monitoring in critical applications like disaster relief. However, limited bandwidth, latency, rain attenuation, long propagation delays, and co-channel interference pose challenges to efficient satellite communication. Therefore, semantic communication (SC) has emerged as a promising solution to improve transmission efficiency and address these issues. In this paper, we explore the potential of SC as a bandwidth-efficient, latency-minimizing strategy specifically suited to 6G satellite communications. While existing SC methods have demonstrated efficacy in direct satellite-terrestrial transmissions, they encounter limitations in satellite networks due to distortion accumulation across gateway hop-relays. Additionally, certain ground users (GUs) experience poor signal-to-noise ratios (SNR), making direct satellite communication challenging. To address these issues, we propose a novel framework that optimizes gateway hop-relay selection for GUs with low SNR and integrates gateway-based denoising mechanisms to ensure high-quality-of-service (QoS) in satellite-based SC networks. This approach directly mitigates distortion, leading to significant improvements in satellite service performance by delivering customized services tailored to the unique signal conditions of each GU. Our findings represent a critical advancement in reliable and efficient data transmission from the Earth observation satellites, thereby enabling fast and effective responses to urgent events. Simulation results demonstrate that our proposed strategy significantly enhances overall network performance, outperforming conventional methods by offering tailored communication services based on specific GU conditions., Comment: 13 pages, 8 figures, 2 tables

Spatial-temporal graph neural networks (STGNNs) have achieved significant success in various time series forecasting tasks. However, due to the lack of explicit and fixed spatial relationships in stock prediction tasks, many STGNNs fail to perform effectively in this domain. While some STGNNs learn spatial relationships from time series, they often lack comprehensiveness. Research indicates that modeling time series using feature changes as tokens reveals entirely different information compared to using time steps as tokens. To more comprehensively extract dynamic spatial information from stock data, we propose a Double-Path Adaptive-correlation Spatial-Temporal Inverted Transformer (DPA-STIFormer). DPA-STIFormer models each node via continuous changes in features as tokens and introduces a Double Direction Self-adaptation Fusion mechanism. This mechanism decomposes node encoding into temporal and feature representations, simultaneously extracting different spatial correlations from a double path approach, and proposes a Double-path gating mechanism to fuse these two types of correlation information. Experiments conducted on four stock market datasets demonstrate state-of-the-art results, validating the model's superior capability in uncovering latent temporal-correlation patterns.

In practical scenarios, federated learning frequently necessitates training personalized models for each client using heterogeneous data. This paper proposes a backbone self-distillation approach to facilitate personalized federated learning. In this approach, each client trains its local model and only sends the backbone weights to the server. These weights are then aggregated to create a global backbone, which is returned to each client for updating. However, the client's local backbone lacks personalization because of the common representation. To solve this problem, each client further performs backbone self-distillation by using the global backbone as a teacher and transferring knowledge to update the local backbone. This process involves learning two components: the shared backbone for common representation and the private head for local personalization, which enables effective global knowledge transfer. Extensive experiments and comparisons with 12 state-of-the-art approaches demonstrate the effectiveness of our approach., Comment: Pubished in ACM MMAsia 2023

A knee point on a curve is the one where the curve levels off after an increase. In a computer system, it marks the point at which the system's performance is no longer improving significantly despite adding extra resources. Thus a knee point often represents an optimal point for decision. However, identifying knee points in noisy data is a challenging task. All previous works defined knee points based on the data in the original scale. However, in this work, we define knee points based on normalized data and provide a mathematical definition of curvature for normalized discrete data points, based on the mathematical definition of curvature for continuous functions. The impact of normalization exerted on curvature and the location of knee points are also discussed. Nevertheless, assessing the effectiveness of methods is difficult in the absence of ground truth data and benchmark datasets, which makes comparing existing methods challenging. In view of this, we create synthetic data that simulate real-world scenarios. We achieve this by selecting a set of functions that possess the required characteristics in this research and then introducing noise that satisfies the underlying distribution. In addition, we present a deep-learning approach and employ a Convolutional Neural Network (CNN) with a U-Net-like architecture, to accurately detect the knee point(s) of the underlying true distribution. The proposed model is evaluated against state-of-the-art methods. Experiments show that our network outperforms existing methods in all synthetic datasets, regardless of whether the samples have single or multiple knee points. In fact, our model achieves the best $F_{1}$ scores among all existing methods in all the test sets.

Diffusion probabilistic models can generate high-quality samples. Yet, their sampling process requires numerous denoising steps, making it slow and computationally intensive. We propose to reduce the sampling cost by pruning a pretrained diffusion model into a mixture of efficient experts. First, we study the similarities between pairs of denoising timesteps, observing a natural clustering, even across different datasets. This suggests that rather than having a single model for all time steps, separate models can serve as ``experts'' for their respective time intervals. As such, we separately fine-tune the pretrained model on each interval, with elastic dimensions in depth and width, to obtain experts specialized in their corresponding denoising interval. To optimize the resource usage between experts, we introduce our Expert Routing Agent, which learns to select a set of proper network configurations. By doing so, our method can allocate the computing budget between the experts in an end-to-end manner without requiring manual heuristics. Finally, with a selected configuration, we fine-tune our pruned experts to obtain our mixture of efficient experts. We demonstrate the effectiveness of our method, DiffPruning, across several datasets, LSUN-Church, LSUN-Beds, FFHQ, and ImageNet, on the Latent Diffusion Model architecture., Comment: Accepted to the 18th European Conference on Computer Vision, ECCV 2024

Confidential Virtual Machines (CVMs) are a type of VMbased Trusted Execution Environments (TEEs) designed to enhance the security of cloud-based VMs, safeguarding them even from malicious hypervisors. Although CVMs have been widely adopted by major cloud service providers, current CVM designs face significant challenges in runtime resource management due to their fixed capacities and lack of transparency. These limitations hamper efficient cloud resource management, leading to increased operational costs and reduced agility in responding to fluctuating workloads. This paper introduces a dynamic CPU resource management approach, featuring the novel concept of "Elastic CVM. This approach allows for hypervisor-assisted runtime adjustment of CPU resources using a specialized vCPU type, termed Worker vCPU. This new approach enhances CPU resource adaptability and operational efficiency without compromising security. Additionally, we introduce a Worker vCPU Abstraction Layer to simplify Worker vCPU deployment and management. To demonstrate the effectiveness of our approach, we have designed and implemented a serverless computing prototype platform, called Ditto. We show that Ditto significantly improves performance and efficiency through finergrain resource management. The concept of "Elastic CVM" and the Worker vCPU design not only optimize cloud resource utilization but also pave the way for more flexible and cost-effective confidential computing environments.

Although current Multi-modal Large Language Models (MLLMs) demonstrate promising results in video understanding, processing extremely long videos remains an ongoing challenge. Typically, MLLMs struggle with handling thousands of tokens that exceed the maximum context length of LLMs, and they experience reduced visual clarity due to token aggregation. Another challenge is the high computational cost stemming from the large number of video tokens. To tackle these issues, we propose Video-XL, an extra-long vision language model designed for efficient hour-scale video understanding. Specifically, we argue that LLMs can be adapted as effective visual condensers and introduce Visual Context Latent Summarization, which condenses visual contexts into highly compact forms. Extensive experiments demonstrate that our model achieves promising results on popular long video understanding benchmarks, despite being trained on limited image data. Moreover, Video-XL strikes a promising balance between efficiency and effectiveness, processing 1024 frames on a single 80GB GPU while achieving nearly 100\% accuracy in the Needle-in-a-Haystack evaluation. We envision Video-XL becoming a valuable tool for long video applications such as video summarization, surveillance anomaly detection, and Ad placement identification.

For fixed $\beta>1$ and $\alpha\in[0,1)$, each $x\in[0,1]$ has an \emph{intermediate $\beta$-expansion} of the form $x=\sum_{i=1}^\infty\frac{c_i-\alpha}{\beta^i}$. Each such expansion produces for the number $x$ a sequence of approximations $\left(\sum_{i=1}^{n}\frac{c_i-\alpha}{\beta^i}\right)_{n\geq1}$. In this paper we study approximation properties by considering $M(\beta,\alpha)$, the expected value of the corresponding \emph{normalized errors} $(\theta_{\beta,\alpha}^n(x))_{n\geq 1}$, given by $$\theta_{\beta,\alpha}^n(x):=\beta^n\left(x-\sum_{i=1}^n\frac{c_i-\alpha}{\beta^i}\right),\quad n\in\mathbb{N}.$$ We prove that $M(\beta,\alpha)$ is continuous in $\beta$ and $\alpha$, respectively. For a fixed $\beta>\sqrt 2$, by the continuity property, $\mathcal{M_\beta}:=\{M(\beta,\alpha):\alpha\in[0,1)\}=[M_\beta,1-M_\beta]$ where $M_\beta=\min\{M(\beta,\alpha):\alpha\in[0,1)\}$. Taking $\beta$ to be a multinacci number, $M(\beta,\alpha)$ is linear on $[0,1-(\beta))$ and locally linear for Lebesgue a.e. $\alpha\in[1-(\beta),1)$; moreover, whether it is increasing or decreasing depends only on the sign of $T_{\beta,\alpha}^{n-1}(1)-T_{\beta,\alpha}^{n-1}(0)$ where $n=\min\{k:T_{\beta,\alpha}^k(1)-T_{\beta,\alpha}^k(0)=0\}$., Comment: 50 page,7 figures

Recent advancements in 3D Gaussian Splatting (3DGS) have unlocked significant potential for modeling 3D head avatars, providing greater flexibility than mesh-based methods and more efficient rendering compared to NeRF-based approaches. Despite these advancements, the creation of controllable 3DGS-based head avatars remains time-intensive, often requiring tens of minutes to hours. To expedite this process, we here introduce the ``Gaussian D\'ej\`a-vu" framework, which first obtains a generalized model of the head avatar and then personalizes the result. The generalized model is trained on large 2D (synthetic and real) image datasets. This model provides a well-initialized 3D Gaussian head that is further refined using a monocular video to achieve the personalized head avatar. For personalizing, we propose learnable expression-aware rectification blendmaps to correct the initial 3D Gaussians, ensuring rapid convergence without the reliance on neural networks. Experiments demonstrate that the proposed method meets its objectives. It outperforms state-of-the-art 3D Gaussian head avatars in terms of photorealistic quality as well as reduces training time consumption to at least a quarter of the existing methods, producing the avatar in minutes., Comment: 11 pages, Accepted by WACV 2025 in Round 1

The fractional quantum Hall effect is a well-known demonstration of strongly correlated topological phases in two dimensions. However, the extension of this phenomenon into a three-dimensional context has yet to be achieved. Recently, the three-dimensional integer quantum Hall effect based on Weyl orbits has been experimentally observed in a topological semimetallic material under a magnetic field. This motivates us to ask whether the Weyl orbits can give rise to the fractional quantum Hall effect when their Landau level is partially filled in the presence of interactions. Here we theoretically demonstrate that the fractional quantum Hall states based on Weyl orbits can emerge in a Weyl semimetal when a Landau level is one-third filled. Using concrete models for Weyl semimetals in magnetic fields, we project the Coulomb interaction onto a single Landau level from the Weyl orbit and find that the ground state of the many-body Hamiltoian is triply degenerate. We further show that the ground states exhibit the many-body Chern number of $1/3$ and the uniform occupation of electrons in both momentum and real space, implying that they are the fractional quantum Hall states. In contrast to the two-dimensional case, the states are spatially localized on two surfaces hosting Fermi arcs. Additionally, our findings suggest that the excitation properties of these states resemble those of the Laughlin state in two dimensions, as inferred from the particle entanglement spectrum of the ground states., Comment: 4 pages, 4figures. Comments are welcome

The properties displayed by graphene at van Hove singularities (VHS) have caught significant attention in recent years. The emergence of exotic quantum states at these singularities prompts investigations on their evolution within the realm of multilayer stacking structures. In our research, we delve into the study of a repulsive Hubbard model focusing on the AA-stacked bilayer graphene at VHS. Within the system's ground state, each of the top and bottom layers hosts a set of spin-density waves (SDWs). These SDWs each takes on three mutually perpendicular spin polarization directions. Importantly, there is noteworthy feature that their spin polarization directions in the two layers exist as elegant embodiments of antiferromagnetic arrangement, persvading the structure with a striking pattern. Referred to in prior research as the chiral SDWs, this intralayer density wave structure confers the system the characteristics of a Chern topological insulator. However, what is particularly fascinating is the pure divergence of the bilayer structure's topological traits when compared to its monolayer counterpart. The system exhibits a profound symmetry known as $Z_2$, preserving its invariance under the combined operations of time-reversal and interlayer exchange. Consequentely, the system's ground state manifests a seemingly trivial Chern number, yet harbors a profound and intricate nontrivial $Z_2$ topological invariant. These remarkable observations align our findings with the conceptual framework of the quantum spin Hall effect.

This paper explores a new method for time series data analysis, aiming to overcome the limitations of traditional mining techniques when dealing with multidimensional time series data. Time series data are extensively utilized in diverse fields, including backend services for monitoring and optimizing IT infrastructure, medical diagnosis through continuous patient monitoring and health trend analysis, and internet business for tracking user behavior and forecasting sales. However, since the effective information in time series data is often hidden in sequence fragments, the uncertainty of their length, quantity, and morphological variables brings challenges to mining. To this end, this paper proposes a new spatiotemporal feature representation method, which converts multidimensional time series (MTS) into one-dimensional event sequences by transforming spatially varying events, and uses a series of event symbols to represent the spatial structural information of multidimensional coupling in the sequence, which has good interpretability. Then, this paper introduces a variable-length tuple mining method to extract non-redundant key event subsequences in event sequences as spatiotemporal structural features of motion sequences. This method is an unsupervised method that does not rely on large-scale training samples and defines a new model for representing the spatiotemporal structural features of multidimensional time series. The superior performance of the STEM model is verified by pattern classification experiments on a variety of motion sequences. The research results of this paper provide an important theoretical basis and technical support for understanding and predicting human behavior patterns, and have far-reaching practical application value.

Improving recommendation systems (RS) can greatly enhance the user experience across many domains, such as social media. Many RS utilize embedding-based retrieval (EBR) approaches to retrieve candidates for recommendation. In an EBR system, the embedding quality is key. According to recent literature, self-supervised multitask learning (SSMTL) has showed strong performance on academic benchmarks in embedding learning and resulted in an overall improvement in multiple downstream tasks, demonstrating a larger resilience to the adverse conditions between each downstream task and thereby increased robustness and task generalization ability through the training objective. However, whether or not the success of SSMTL in academia as a robust training objectives translates to large-scale (i.e., over hundreds of million users and interactions in-between) industrial RS still requires verification. Simply adopting academic setups in industrial RS might entail two issues. Firstly, many self-supervised objectives require data augmentations (e.g., embedding masking/corruption) over a large portion of users and items, which is prohibitively expensive in industrial RS. Furthermore, some self-supervised objectives might not align with the recommendation task, which might lead to redundant computational overheads or negative transfer. In light of these two challenges, we evaluate using a robust training objective, specifically SSMTL, through a large-scale friend recommendation system on a social media platform in the tech sector, identifying whether this increase in robustness can work at scale in enhancing retrieval in the production setting. Through online A/B testing with SSMTL-based EBR, we observe statistically significant increases in key metrics in the friend recommendations, with up to 5.45% improvements in new friends made and 1.91% improvements in new friends made with cold-start users., Comment: RobustRecSys workshop @ RecSys 2024

One type of bond with the most implicit government guarantee is municipal investment bonds. In recent years, there have been an increasing number of downgrades in the credit ratings of municipal bonds, which has led some people to question whether the implicit government guarantee may affect the objectivity of the bond ratings? This paper uses text mining methods to mine relevant policy documents related to municipal investment bond issuance, and calculates the implicit guarantee strength of municipal investment bonds based on the PMC index model. It further analyzes the impact of the implicit guarantee strength of municipal bonds on their credit evaluation. The study found that the implicit government guarantee on municipal investment bonds does indeed help to raise the credit ratings assigned by credit rating agencies. The study found that, moreover, the government's implicit guarantee has a more pronounced effect in boosting credit ratings in less developed western regions., Comment: 16pages,1 figure

Obtaining word timestamp information from end-to-end (E2E) ASR models remains challenging due to the lack of explicit time alignment during training. This issue is further complicated in multilingual models. Existing methods, either rely on lexicons or introduce additional tokens, leading to scalability issues and increased computational costs. In this work, we propose a new approach to estimate word boundaries without relying on lexicons. Our method leverages word embeddings from sub-word token units and a pretrained ASR model, requiring only word alignment information during training. Our proposed method can scale-up to any number of languages without incurring any additional cost. We validate our approach using a multilingual ASR model trained on five languages and demonstrate its effectiveness against a strong baseline., Comment: Submitted to ICASSP 2025

Speech enhancement (SE) aims to extract the clean waveform from noise-contaminated measurements to improve the speech quality and intelligibility. Although learning-based methods can perform much better than traditional counterparts, the large computational complexity and model size heavily limit the deployment on latency-sensitive and low-resource edge devices. In this work, we propose a lightweight SE network (LiSenNet) for real-time applications. We design sub-band downsampling and upsampling blocks and a dual-path recurrent module to capture band-aware features and time-frequency patterns, respectively. A noise detector is developed to detect noisy regions in order to perform SE adaptively and save computational costs. Compared to recent higher-resource-dependent baseline models, the proposed LiSenNet can achieve a competitive performance with only 37k parameters (half of the state-of-the-art model) and 56M multiply-accumulate (MAC) operations per second., Comment: 5 pages, submitted to 2025 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2025)

The residual stress greatly affects the mechanical behavior of a material. In this work, the effect of residual stress on the isothermal tensile behavior of a NiTi shape memory alloy is studied. The focused ion beam and digital image correlation are combined to measure the two-dimensional residual stress in nanocrystalline NiTi plates processed with prestrain laser shock peening. A four-point bending experiment verified the accuracy of this measurement method. The FIB-DIC method is an attractive tool for measuring the two-dimensional residual stress in phase transition nanocrystalline materials. The internal residual stress significantly decreases the phase transition stress, and the mechanism is studied via finite element and theoretical analyses. This work implies that the mechanical behavior of NiTi shape memory alloys can be tailored via residual stress engineering., Comment: 43 pages, 20 figures

Magnetic topological insulators exhibit unique electronic states due to the interplay between the electronic topology and the spin structure. The antiferromagnetic metal $\rm{EuIn_2As_2}$ is a prominent candidate material in which exotic topological phases, including an axion insulating state, are theoretically predicted depending on the magnetic structure of the $\rm{Eu}^{2+}$ moments. Here, we report experimental results of the nuclear magnetic resonance (NMR) measurements of all the nuclei in $\rm{EuIn_2As_2}$ to investigate the coupling between the magnetic moments in the Eu ions and the conduction electrons in $\rm{In_2As_2}$ layers and the magnetic structure. The $^{75}\rm{As}$ and $^{115}\rm{In}$ NMR spectra observed at zero external magnetic field reveal the appearance of internal fields of $4.9$ and $3.6\ \rm{T}$ respectively at the lowest temperature, suggesting a strong coupling between the conduction electrons in the $\rm{In_2As_2}$ layer and the ordered magnetic moments in the Eu ions. The $^{75}\rm{As}$ NMR spectra under in-plane external magnetic fields show broad distributions of the internal fields produced by an incommensurate fan-like spin structure which turns into a forced ferromagnetic state above $0.7\ \rm{T}$. We propose a spin reorientation process that an incommensurate helical state at zero external magnetic field quickly changes into a fan state by applying a slight magnetic field., Comment: 19 pages, 6 figures

Navigating and understanding complex environments over extended periods of time is a significant challenge for robots. People interacting with the robot may want to ask questions like where something happened, when it occurred, or how long ago it took place, which would require the robot to reason over a long history of their deployment. To address this problem, we introduce a Retrieval-augmented Memory for Embodied Robots, or ReMEmbR, a system designed for long-horizon video question answering for robot navigation. To evaluate ReMEmbR, we introduce the NaVQA dataset where we annotate spatial, temporal, and descriptive questions to long-horizon robot navigation videos. ReMEmbR employs a structured approach involving a memory building and a querying phase, leveraging temporal information, spatial information, and images to efficiently handle continuously growing robot histories. Our experiments demonstrate that ReMEmbR outperforms LLM and VLM baselines, allowing ReMEmbR to achieve effective long-horizon reasoning with low latency. Additionally, we deploy ReMEmbR on a robot and show that our approach can handle diverse queries. The dataset, code, videos, and other material can be found at the following link: https://nvidia-ai-iot.github.io/remembr

The rapid development of Large Vision-Language Models (LVLMs) often comes with widespread hallucination issues, making cost-effective and comprehensive assessments increasingly vital. Current approaches mainly rely on costly annotations and are not comprehensive -- in terms of evaluating all aspects such as relations, attributes, and dependencies between aspects. Therefore, we introduce the FIHA (autonomous Fine-graIned Hallucination evAluation evaluation in LVLMs), which could access hallucination LVLMs in the LLM-free and annotation-free way and model the dependency between different types of hallucinations. FIHA can generate Q&A pairs on any image dataset at minimal cost, enabling hallucination assessment from both image and caption. Based on this approach, we introduce a benchmark called FIHA-v1, which consists of diverse questions on various images from MSCOCO and Foggy. Furthermore, we use the Davidson Scene Graph (DSG) to organize the structure among Q&A pairs, in which we can increase the reliability of the evaluation. We evaluate representative models using FIHA-v1, highlighting their limitations and challenges. We released our code and data.

The surge of hate speech on social media platforms poses a significant challenge, with hate speech detection~(HSD) becoming increasingly critical. Current HSD methods focus on enriching contextual information to enhance detection performance, but they overlook the inherent uncertainty of hate speech. We propose a novel HSD method, named trustworthy hate speech detection method through visual augmentation (TrusV-HSD), which enhances semantic information through integration with diffused visual images and mitigates uncertainty with trustworthy loss. TrusV-HSD learns semantic representations by effectively extracting trustworthy information through multi-modal connections without paired data. Our experiments on public HSD datasets demonstrate the effectiveness of TrusV-HSD, showing remarkable improvements over conventional methods.

Data wrangling, the process of preparing raw data for further analysis in computational notebooks, is a crucial yet time-consuming step in data science. Code generation has the potential to automate the data wrangling process to reduce analysts' overhead by translating user intents into executable code. Precisely generating data wrangling code necessitates a comprehensive consideration of the rich context present in notebooks, including textual context, code context and data context. However, notebooks often interleave multiple non-linear analysis tasks into linear sequence of code blocks, where the contextual dependencies are not clearly reflected. Directly training models with source code blocks fails to fully exploit the contexts for accurate wrangling code generation. To bridge the gap, we aim to construct a high quality datasets with clear and rich contexts to help training models for data wrangling code generation tasks. In this work, we first propose an automated approach, CoCoMine to mine data-wrangling code generation examples with clear multi-modal contextual dependency. It first adopts data flow analysis to identify the code blocks containing data wrangling codes. Then, CoCoMine extracts the contextualized datawrangling code examples through tracing and replaying notebooks. With CoCoMine, we construct CoCoNote, a dataset containing 58,221 examples for Contextualized Data-wrangling Code generation in Notebooks. To demonstrate the effectiveness of our dataset, we finetune a range of pretrained code models and prompt various large language models on our task. Furthermore, we also propose DataCoder, which encodes data context and code&textual contexts separately to enhance code generation. Experiment results demonstrate the significance of incorporating data context in data-wrangling code generation and the effectiveness of our model. We release code and data at url..., Comment: To appear at ASE 2024

Rutile RuO$_2$ has been posited as a potential $d$-wave altermagnetism candidate, with a predicted significant spin splitting up to 1.4 eV. Despite accumulating theoretical predictions and transport measurements, direct spectroscopic observation of spin splitting has remained elusive. Here, we employ spin- and angle-resolved photoemission spectroscopy to investigate the band structures and spin polarization of thin-film and single-crystal RuO$_2$. Contrary to expectations of altermagnetism, our analysis indicates that RuO$_2$'s electronic structure aligns with those predicted under non-magnetic conditions, exhibiting no evidence of the hypothesized spin splitting. Additionally, we observe significant in-plane spin polarization of the low-lying bulk bands, which is antisymmetric about the high-symmetry plane and contrary to the $d$-wave spin texture due to time-reversal symmetry breaking in altermagnetism. These findings definitively challenge the altermagnetic order previously proposed for rutile RuO$_2$, prompting a reevaluation of its magnetic properties., Comment: 6 pages, 4 figures. Submitted to Physical Review Letters

Angular momentum (AM), a fundamental concept describing the rotation of an object about an axis, profoundly influences all branches of physics. In condensed matter, AM is intimately related to the emergence of topological quantum states, including chiral superconductivity and quantum spin liquids, and various chiral quasiparticles. Recently, it has been predicted that microscopic lattice excitations, known as phonons, can carry finite AM with remarkable macroscopic physical consequences. However, the direct observation of phonon-AM has not been achieved. In this letter, we report the experimental discovery of phonon-AM in the chiral crystal Tellurium. We show that due to AM conservation, applying a time-reversal symmetry breaking thermal gradient along the chiral axis of single crystal Te results in a macroscopic mechanical torque, $\tau$, that can be observed using a cantilever-based device. We establish that the mechanical torques change sign by flipping the thermal gradient and disappear in polycrystalline samples that lack a preferred chirality. Based on our experimental settings, we estimate $\tau\sim10^{-11}$N$\cdot$m, in agreement with theoretical calculations. Our results uncover phonon-AM and pave the way for phonon-AM enabled quantum states for microelectronic applications.

With the terminal value $|\xi|$ admitting some given exponential moment, we put forward and prove several existence and uniqueness results for the unbounded solutions of quadratic backward stochastic differential equations whose generators may be represented as a uniformly continuous (not necessarily locally Lipschitz continuous) perturbation of some convex (concave) function with quadratic growth. These results generalize those posed in \cite{Delbaen 2011} and \cite{Fan-Hu-Tang 2020} to some extent. The critical case is also tackled, which strengthens the main result of \cite{Delbaen 2015}., Comment: 22 pages

Reducing latency is a roaring trend in recent super-resolution (SR) research. While recent progress exploits various convolutional blocks, attention modules, and backbones to unlock the full potentials of the convolutional neural network (ConvNet), achieving real-time performance remains a challenge. To this end, we present PlainUSR, a novel framework incorporating three pertinent modifications to expedite ConvNet for efficient SR. For the convolutional block, we squeeze the lighter but slower MobileNetv3 block into a heavier but faster vanilla convolution by reparameterization tricks to balance memory access and calculations. For the attention module, by modulating input with a regional importance map and gate, we introduce local importance-based attention to realize high-order information interaction within a 1-order attention latency. As to the backbone, we propose a plain U-Net that executes channel-wise discriminate splitting and concatenation. In the experimental phase, PlainUSR exhibits impressively low latency, great scalability, and competitive performance compared to both state-of-the-art latency-oriented and quality-oriented methods. In particular, compared to recent NGswin, the PlainUSR-L is 16.4x faster with competitive performance., Comment: Accepted by ACCV 2024. Under camera-ready revision

Massive communication is one of key scenarios of 6G where two magnitude higher connection density would be required to serve diverse services. As a promising direction, unsourced multiple access has been proved to outperform significantly over orthogonal multiple access (OMA) or slotted-ALOHA in massive connections. In this paper we describe a design framework of unsourced sparse multiple access (USMA) that consists of two key modules: compressed sensing for preamble generation, and sparse interleaver division multiple access (SIDMA) for main packet transmission. Simulation results of general design of USMA show that the theoretical bound can be approached within 1~1.5 dB by using simple channel codes like convolutional. To illustrate the scalability of USMA, a customized design for ambient Internet of Things (A-IoT) is proposed, so that much less memory and computation are required. Simulations results of Rayleigh fading and realistic channel estimation show that USMA based A-IoT solution can deliver nearly 4 times capacity and 6 times efficiency for random access over traditional radio frequency identification (RFID) technology., Comment: 7 pages, 5 figures and 1 table

Emotional intelligence in large language models (LLMs) is of great importance in Natural Language Processing. However, the previous research mainly focus on basic sentiment analysis tasks, such as emotion recognition, which is not enough to evaluate LLMs' overall emotional intelligence. Therefore, this paper presents a novel framework named EmotionQueen for evaluating the emotional intelligence of LLMs. The framework includes four distinctive tasks: Key Event Recognition, Mixed Event Recognition, Implicit Emotional Recognition, and Intention Recognition. LLMs are requested to recognize important event or implicit emotions and generate empathetic response. We also design two metrics to evaluate LLMs' capabilities in recognition and response for emotion-related statements. Experiments yield significant conclusions about LLMs' capabilities and limitations in emotion intelligence., Comment: Accepted to ACL 2024 (Findings)

Gramian-based model order reduction methods, like balanced truncation, and interpolation-based methods, such as H2-optimal reduction, are two important types of model reduction algorithms. Although both are known for their accuracy, they are often seen as two different approaches. This paper shows that these two methods are closely related, with Gramian-based reduction being roughly an interpolation problem, and vice versa. For Galerkin projection, we find that when the reduced model has enough order to capture the significant eigenvalues of the controllability or observability Gramian, preserving these eigenvalues becomes an interpolation problem. In this case, both Gramian-based and interpolation-based model reduction methods produce the same transfer function but with different state-space realizations. When the reduced model's order is too small to capture all significant eigenvalues, the methods begin to differ, and the difference depends on the eigenvalues that were left out. In the case of Petrov-Galerkin projection, if the reduced model's order is large enough to capture the significant Hankel singular values, balanced truncation becomes the same as H2-optimal model order reduction. Again, both methods give the same transfer function but with different state-space realizations. When the order is smaller, the methods diverge, with the difference depending on the truncated Hankel singular values. Numerical examples are provided to support these findings, showing that Gramian-based and interpolation-based model reduction methods are more connected than previously thought and can be viewed as approximations of each other.

The integration of hardware accelerators has significantly advanced the capabilities of modern recommendation systems, enabling the exploration of complex ranking paradigms previously deemed impractical. However, the GPU-based computational costs present substantial challenges. In this paper, we demonstrate our development of an efficiency-driven approach to explore these paradigms, moving beyond traditional reliance on native PyTorch modules. We address the specific challenges posed by ranking models' dependence on categorical features, which vary in length and complicate GPU utilization. We introduce Jagged Feature Interaction Kernels, a novel method designed to extract fine-grained insights from long categorical features through efficient handling of dynamically sized tensors. We further enhance the performance of attention mechanisms by integrating Jagged tensors with Flash Attention. Our novel Jagged Flash Attention achieves up to 9x speedup and 22x memory reduction compared to dense attention. Notably, it also outperforms dense flash attention, with up to 3x speedup and 53% more memory efficiency. In production models, we observe 10% QPS improvement and 18% memory savings, enabling us to scale our recommendation systems with longer features and more complex architectures., Comment: 3 pages, 2 figures

gVisor is a Google-published application-level kernel for containers. As gVisor is lightweight and has sound isolation, it has been widely used in many IT enterprises \cite{Stripe, DigitalOcean, Cloundflare}. When a new vulnerability of the upstream gVisor is found, it is important for the downstream developers to test the corresponding code to maintain the security. To achieve this aim, directed fuzzing is promising. Nevertheless, there are many challenges in applying existing directed fuzzing methods for gVisor. The core reason is that existing directed fuzzers are mainly for general C/C++ applications, while gVisor is an OS kernel written in the Go language. To address the above challenges, we propose G-Fuzz, a directed fuzzing framework for gVisor. There are three core methods in G-Fuzz, including lightweight and fine-grained distance calculation, target related syscall inference and utilization, and exploration and exploitation dynamic switch. Note that the methods of G-Fuzz are general and can be transferred to other OS kernels. We conduct extensive experiments to evaluate the performance of G-Fuzz. Compared to Syzkaller, the state-of-the-art kernel fuzzer, G-Fuzz outperforms it significantly. Furthermore, we have rigorously evaluated the importance for each core method of G-Fuzz. G-Fuzz has been deployed in industry and has detected multiple serious vulnerabilities., Comment: This paper has published in IEEE Transactions on Dependable and Secure Computing (TDSC), https://ieeexplore.ieee.org/abstract/document/10049484/citations?tabFilter=papers#citations

Real-time SLAM with dense 3D mapping is computationally challenging, especially on resource-limited devices. The recent development of 3D Gaussian Splatting (3DGS) offers a promising approach for real-time dense 3D reconstruction. However, existing 3DGS-based SLAM systems struggle to balance hardware simplicity, speed, and map quality. Most systems excel in one or two of the aforementioned aspects but rarely achieve all. A key issue is the difficulty of initializing 3D Gaussians while concurrently conducting SLAM. To address these challenges, we present Monocular GSO (MGSO), a novel real-time SLAM system that integrates photometric SLAM with 3DGS. Photometric SLAM provides dense structured point clouds for 3DGS initialization, accelerating optimization and producing more efficient maps with fewer Gaussians. As a result, experiments show that our system generates reconstructions with a balance of quality, memory efficiency, and speed that outperforms the state-of-the-art. Furthermore, our system achieves all results using RGB inputs. We evaluate the Replica, TUM-RGBD, and EuRoC datasets against current live dense reconstruction systems. Not only do we surpass contemporary systems, but experiments also show that we maintain our performance on laptop hardware, making it a practical solution for robotics, A/R, and other real-time applications., Comment: Paper Contribution to the ICRA 2025 Conference. Currently being reviewed

Advancements in Large Language Models (LLMs) inspire various strategies for integrating video modalities. A key approach is Video-LLMs, which incorporate an optimizable interface linking sophisticated video encoders to LLMs. However, due to computation and data limitations, these Video-LLMs are typically pre-trained to process only short videos, limiting their broader application for understanding longer video content. Additionally, fine-tuning Video-LLMs to handle longer videos is cost-prohibitive. Consequently, it becomes essential to explore the interpolation of Video-LLMs under a completely training-free setting. In this paper, we first identify the primary challenges in interpolating Video-LLMs: (1) the video encoder and modality alignment projector are fixed, preventing the integration of additional frames into Video-LLMs, and (2) the LLM backbone is limited in its content length capabilities, which complicates the processing of an increased number of video tokens. To address these challenges, we propose a specific INTerPolation method for Video-LLMs (INTP-Video-LLMs). We introduce an alternative video token rearrangement technique that circumvents limitations imposed by the fixed video encoder and alignment projector. Furthermore, we introduce a training-free LLM context window extension method to enable Video-LLMs to understand a correspondingly increased number of visual tokens.

In this paper, we present a novel approach, called LogicPro, to enhance Large Language Models (LLMs) complex Logical reasoning through Program Examples. We do this effectively by simply utilizing widely available algorithmic problems and their code solutions. First, we constructed diverse test samples input based on algorithmic questions and code solutions. Then, we designed different complex reasoning questions based on algorithmic problems and test samples. Finally, combining the intermediate variable outputs of the code solutions and the complex reasoning questions, we derived the reasoning process and the final answer. With this approach, we can construct a dataset that is sufficiently difficult (all models are ineffective), diverse (synthesized from 2,360 different algorithmic questions), and scalable (building different test samples and collecting more algorithmic questions). In addition, we obtain a high-quality reasoning process guided by the values of intermediate variables. As a result, our approach achieves significant improvements in multiple models for the BBH$^{27}$, GSM8K, HellSwag, Logicqa, Reclor, and RTE datasets, outperforming a wide range of existing reasoning datasets.

The implicit government guarantee hampers the recognition and management of risks by all stakeholders in the bond market, and it has led to excessive debt for local governments or state-owned enterprises. To prevent the risk of local government debt defaults and reduce investors' expectations of implicit government guarantees, various regulatory departments have issued a series of policy documents related to municipal investment bonds. By employing text mining techniques on policy documents related to municipal investment bond, and utilizing the PMC index model to assess the effectiveness of policy documents. This paper proposes a novel method for quantifying the intensity of implicit governmental guarantees based on PMC index model. The intensity of implicit governmental guarantees is inversely correlated with the PMC index of policies aimed at de-implicitizing governmental guarantees. Then as these policies become more effective, the intensity of implicit governmental guarantees diminishes correspondingly. These findings indicate that recent policies related to municipal investment bond have indeed succeeded in reducing implicit governmental guarantee intensity, and these policies have achieved the goal of risk management. Furthermore, it was showed that the intensity of implicit governmental guarantee affected by diverse aspects of these policies such as effectiveness, clarity, and specificity, as well as incentive and assurance mechanisms., Comment: 22 pages,6 figures

This paper is concerned with the reconstruction of the shape of an acoustic obstacle. Based on the use of the tapered waves with very narrow widths illuminating the obstacle, the boundary of the obstacle is reconstructed by a direct imaging algorithm. The stability of the imaging scheme is mathematically analyzed. We emphasize that different from the incident plane waves or point sources, the tapered waves with narrow widths bring several benefits in the inverse scattering: 1. local property. A tapered wave can illuminate only on a local part of the boundary of the obstacle, which generates the scattered field; 2. high resolution. We need only reconstruct the boundary near the beam, which improves the quality of some well-known algorithms; 3. fast and easy to implement. Numerical examples are included to demonstrate the effectiveness of the tapered waves., Comment: 17 pages, 11 figures

Powder bed fusion (PBF) is an emerging metal additive manufacturing (AM) technology that enables rapid fabrication of complex geometries. However, defects such as pores and balling may occur and lead to structural unconformities, thus compromising the mechanical performance of the part. This has become a critical challenge for quality assurance as the nature of some defects is stochastic during the process and invisible from the exterior. To address this issue, digital twin (DT) using machine learning (ML)-based modeling can be deployed for AM process monitoring and control. Melt pool is one of the most commonly observed physical phenomena for process monitoring, usually by high-speed cameras. Once labeled and preprocessed, the melt pool images are used to train ML-based models for DT applications such as process anomaly detection and print quality evaluation. Nonetheless, the reusability of DTs is restricted due to the wide variability of AM settings, including AM machines and monitoring instruments. The performance of the ML models trained using the dataset collected from one setting is usually compromised when applied to other settings. This paper proposes a knowledge transfer pipeline between different AM settings to enhance the reusability of AM DTs. The source and target datasets are collected from the National Institute of Standards and Technology and National Cheng Kung University with different cameras, materials, AM machines, and process parameters. The proposed pipeline consists of four steps: data preprocessing, data augmentation, domain alignment, and decision alignment. Compared with the model trained only using the source dataset, this pipeline increased the melt pool anomaly detection accuracy by 31% without any labeled training data from the target dataset., Comment: 8 pages, 7 figures, 3 tables. IEEE CASE 2024

The increasing use of text as data in social science research necessitates the development of valid, consistent, reproducible, and efficient methods for generating text-based concept measures. This paper presents a novel method that leverages the internal hidden states of large language models (LLMs) to generate these concept measures. Specifically, the proposed method learns a concept vector that captures how the LLM internally represents the target concept, then estimates the concept value for text data by projecting the text's LLM hidden states onto the concept vector. Three replication studies demonstrate the method's effectiveness in producing highly valid, consistent, and reproducible text-based measures across various social science research contexts, highlighting its potential as a valuable tool for the research community.

Quantum collision model provides a promising tool for investigating system-bath dynamics. Most of the studies on quantum collision models work in the resonant regime. In quantum dynamics, the off-resonant interaction often brings in exciting ffects. It is thereby attractive to investigate quantum collision models in the off-resonant regime. On the other hand, a bath with a negative absolute temperature is anticipated to be instrumental in developing thermal devices. The design of an effective bath with negative absolute temperature coupled to a qubit is significant for developing such thermal devices. We establish an effective negative-absolute-temperature bath coupled to a qubit with a quantum collision model in a far-off-resonant regime. We conduct a detailed and systematic investigation on the off-resonant collision model. There is an additional constraint on the collision duration resulting from the far-off resonant collision. The dynamics of the collision model in the far-off-resonant regime are different from the one beyond the far-off-resonant regime. Numerical simulations confirm the validity of the proposed approach., Comment: 7 pages, 3 figures

High-performance programmable silicon photonic circuits are considered to be a critical part of next generation architectures for optical processing, photonic quantum circuits and neural networks. Low-loss optical phase change materials (PCMs) offer a promising route towards non-volatile free-form control of light. Here, we exploit direct-write digital patterning of waveguides using layers of the PCM Sb$_2$Se$_3$ with a thickness of up to 100 nm, demonstrating the ability to strongly increase the effect per pixel compared to previous implementations where much thinner PCM layers were used. We exploit the excellent refractive index matching between Sb$_2$Se$_3$ and silicon to achieve a low-loss hybrid platform for programmable photonics. A five-fold reduction in modulation length of a Mach-Zehnder interferometer is achieved compared to previous work using thin-film Sb$_2$Se$_3$ devices, decreased to 5 $\mu$m in this work. Application of the thicker PCM layers in direct-write digital programming of a multimode interferometer (MMI) shows a three-fold reduction of the number of programmed pixels to below 10 pixels per device. The demonstrated scaling of performance with PCM layer thickness is important for establishing the optimum working range for hybrid silicon-PCM devices and holds promise for achieving ultracompact programmable photonic circuits., Comment: 24 pages, 5 figures

Video generation using diffusion-based models is constrained by high computational costs due to the frame-wise iterative diffusion process. This work presents a Diffusion Reuse MOtion (Dr. Mo) network to accelerate latent video generation. Our key discovery is that coarse-grained noises in earlier denoising steps have demonstrated high motion consistency across consecutive video frames. Following this observation, Dr. Mo propagates those coarse-grained noises onto the next frame by incorporating carefully designed, lightweight inter-frame motions, eliminating massive computational redundancy in frame-wise diffusion models. The more sensitive and fine-grained noises are still acquired via later denoising steps, which can be essential to retain visual qualities. As such, deciding which intermediate steps should switch from motion-based propagations to denoising can be a crucial problem and a key tradeoff between efficiency and quality. Dr. Mo employs a meta-network named Denoising Step Selector (DSS) to dynamically determine desirable intermediate steps across video frames. Extensive evaluations on video generation and editing tasks have shown that Dr. Mo can substantially accelerate diffusion models in video tasks with improved visual qualities.

Digital twin (DT) technology has a high potential to satisfy different requirements of the ever-expanding new applications. Nonetheless, the DT placement in wireless digital twin networks (WDTNs) poses a significant challenge due to the conflict between unpredictable workloads and the limited capacity of edge servers. In other words, each edge server has a risk of overload when handling an excessive number of tasks or services. Overload risks can have detrimental effects on a network's sustainability, yet this aspect is often overlooked in the literature. In this paper, we aim to study the sustainability-aware DT placement problem for WDTNs from a cost minimization perspective. To this end, we formulate the DT placement-driven cost optimization problem as a chance-constrained integer programming problem. For tractability, we transform the original non-deterministic problem into a deterministic integer linear programming (ILP) problem using the sample average approximation (SAA) approach. We prove that the transformed problem remains NP-hard and thus finding a global optimal solution is very difficult. To strike a balance between time efficiency and performance guarantee, we propose an improved local search algorithm for this ILP by identifying high-quality starting states from historical search data and enhancing the search process. Numerical results show a lower cost and higher efficiency of our proposed method compared with the previous schemes.

Inspired by the gradient flow viewpoint of the Landau equation and corresponding dynamic formulation of the Landau metric in [arXiv:2007.08591], we develop a novel implicit particle method for the Landau equation in the framework of the JKO scheme. We first reformulate the Landau metric in a computationally friendly form, and then translate it into the Lagrangian viewpoint using the flow map. A key observation is that, while the flow map evolves according to a rather complicated integral equation, the unknown component is merely a score function of the corresponding density plus an additional term in the null space of the collision kernel. This insight guides us in approximating the flow map with a neural network and simplifies the training. Additionally, the objective function is in a double summation form, making it highly suitable for stochastic methods. Consequently, we design a tailored version of stochastic gradient descent that maintains particle interactions and reduces the computational complexity. Compared to other deterministic particle methods, the proposed method enjoys exact entropy dissipation and unconditional stability, therefore making it suitable for large-scale plasma simulations over extended time periods.

Broad textual understanding and in-context learning require language models that utilize full document contexts. Due to the implementation challenges associated with directly training long-context models, many methods have been proposed for extending models to handle long contexts. However, owing to differences in data and model classes, it has been challenging to compare these approaches, leading to uncertainty as to how to evaluate long-context performance and whether it differs from standard evaluation. We implement a controlled protocol for extension methods with a standardized evaluation, utilizing consistent base models and extension data. Our study yields several insights into long-context behavior. First, we reaffirm the critical role of perplexity as a general-purpose performance indicator even in longer-context tasks. Second, we find that current approximate attention methods systematically underperform across long-context tasks. Finally, we confirm that exact fine-tuning based methods are generally effective within the range of their extension, whereas extrapolation remains challenging. All codebases, models, and checkpoints will be made available open-source, promoting transparency and facilitating further research in this critical area of AI development.

With the advent of the big data and large language model era, zero-shot personalized rapid customization has emerged as a significant trend. In this report, we introduce Takin AudioLLM, a series of techniques and models, mainly including Takin TTS, Takin VC, and Takin Morphing, specifically designed for audiobook production. These models are capable of zero-shot speech production, generating high-quality speech that is nearly indistinguishable from real human speech and facilitating individuals to customize the speech content according to their own needs. Specifically, we first introduce Takin TTS, a neural codec language model that builds upon an enhanced neural speech codec and a multi-task training framework, capable of generating high-fidelity natural speech in a zero-shot way. For Takin VC, we advocate an effective content and timbre joint modeling approach to improve the speaker similarity, while advocating for a conditional flow matching based decoder to further enhance its naturalness and expressiveness. Last, we propose the Takin Morphing system with highly decoupled and advanced timbre and prosody modeling approaches, which enables individuals to customize speech production with their preferred timbre and prosody in a precise and controllable manner. Extensive experiments validate the effectiveness and robustness of our Takin AudioLLM series models. For detailed demos, please refer to https://everest-ai.github.io/takinaudiollm/., Comment: Technical Report; 18 pages; typos corrected, references added, demo url modified, author name modified