arXiv subjects

Kai Wang

Publications and source records attributed to Kai Wang.

At least 451 records · Page 25Linked to original sources

arXiv2024

Number rigid determinantal point processes induced by generalized Cantor sets

We consider the Ghosh-Peres number rigidity of translation-invariant determinantal point processes on the real line $\mathbb{R}$, whose correlation kernels are induced by the Fourier transform of the indicators of generalized Cantor sets in the unit interval. Our main results show that for any given $θ\in(0,1)$, there exists a generalized Cantor set with Lebesgue measure $θ$, such that the corresponding determinantal point process is Ghosh-Peres number rigid.

math.PR↗

arXiv2024

Advancing Large Multi-modal Models with Explicit Chain-of-Reasoning and Visual Question Generation

The increasing demand for intelligent systems capable of interpreting and reasoning about visual content requires the development of large Vision-and-Language Models (VLMs) that are not only accurate but also have explicit reasoning capabilities. This paper presents a novel approach to develop a VLM with the ability to conduct explicit reasoning based on visual content and textual instructions. We introduce a system that can ask a question to acquire necessary knowledge, thereby enhancing the robustness and explicability of the reasoning process. To this end, we developed a novel dataset generated by a Large Language Model (LLM), designed to promote chain-of-thought reasoning combined with a question-asking mechanism. The dataset covers a range of tasks, from common ones like caption generation to specialized VQA tasks that require expert knowledge. Furthermore, using the dataset we created, we fine-tuned an existing VLM. This training enabled the models to generate questions and perform iterative reasoning during inference. The results demonstrated a stride toward a more robust, accurate, and interpretable VLM, capable of reasoning explicitly and seeking information proactively when confronted with ambiguous visual input.

cs.CV↗

arXiv2024

High-energy neutrino emission from tidal disruption event outflow-cloud interactions

Tidal disruption events (TDEs), characterized by their luminous transients and high-velocity outflows, have emerged as plausible sources of high-energy neutrinos contributing to the diffuse neutrino. In this study, we calculate the contribution of TDEs to the diffuse neutrino by employing the outflow-cloud model within the TDE framework. Our analysis indicates that the contribution of TDEs becomes negligible when the redshift $Z$ exceeds 2. Employing a set of fiducial values, which includes outflow energy $E_{\rm kin}=10^{51}$ erg, a proton spectrum cutoff energy $E_{\rm p,max}=100$ PeV, a volume TDE rate $\dot{N}=8 \times 10^{-7}\ \rm Mpc^{-3}\ year^{-1}$, covering fraction of clouds $C_V=0.1$, energy conversion efficiency in the shock $η=0.1$, and a proton spectrum index $Γ=-1.7$, we find that TDEs can account for approximately 80\% of the contribution at energies around 0.3 PeV. Additionally, TDEs still contribute around 18\% to the IceCube data below 0.1 PeV and the total contribution is $\sim 24^{+2}_{-15}\%$. In addition, we also discuss the potential influence of various parameter values on the results in detail. With the IceCube data, we impose constraints on the combination of the physical parameters, i.e., $C_{f}=\dot{N}E_{\rm kin}C_{\rm v}η$. Future observations or theoretical considerations would fix some physical parameters, which will help to constrain some individual parameters of TDEs.

astro-ph.HE↗

arXiv2024

Exemplar-free Continual Representation Learning via Learnable Drift Compensation

Exemplar-free class-incremental learning using a backbone trained from scratch and starting from a small first task presents a significant challenge for continual representation learning. Prototype-based approaches, when continually updated, face the critical issue of semantic drift due to which the old class prototypes drift to different positions in the new feature space. Through an analysis of prototype-based continual learning, we show that forgetting is not due to diminished discriminative power of the feature extractor, and can potentially be corrected by drift compensation. To address this, we propose Learnable Drift Compensation (LDC), which can effectively mitigate drift in any moving backbone, whether supervised or unsupervised. LDC is fast and straightforward to integrate on top of existing continual learning approaches. Furthermore, we showcase how LDC can be applied in combination with self-supervised CL methods, resulting in the first exemplar-free semi-supervised continual learning approach. We achieve state-of-the-art performance in both supervised and semi-supervised settings across multiple datasets. Code is available at \url{https://github.com/alviur/ldc}.

cs.CV↗

arXiv2024

ColorPeel: Color Prompt Learning with Diffusion Models via Color and Shape Disentanglement

Text-to-Image (T2I) generation has made significant advancements with the advent of diffusion models. These models exhibit remarkable abilities to produce images based on textual prompts. Current T2I models allow users to specify object colors using linguistic color names. However, these labels encompass broad color ranges, making it difficult to achieve precise color matching. To tackle this challenging task, named color prompt learning, we propose to learn specific color prompts tailored to user-selected colors. Existing T2I personalization methods tend to result in color-shape entanglement. To overcome this, we generate several basic geometric objects in the target color, allowing for color and shape disentanglement during the color prompt learning. Our method, denoted as ColorPeel, successfully assists the T2I models to peel off the novel color prompts from these colored shapes. In the experiments, we demonstrate the efficacy of ColorPeel in achieving precise color generation with T2I models. Furthermore, we generalize ColorPeel to effectively learn abstract attribute concepts, including textures, materials, etc. Our findings represent a significant step towards improving precision and versatility of T2I models, offering new opportunities for creative applications and design tasks. Our project is available at https://moatifbutt.github.io/colorpeel/.

cs.CV↗

arXiv2024

AViTMP: A Tracking-Specific Transformer for Single-Branch Visual Tracking

Visual object tracking is a fundamental component of transportation systems, especially for intelligent driving. Despite achieving state-of-the-art performance in visual tracking, recent single-branch trackers tend to overlook the weak prior assumptions associated with the Vision Transformer (ViT) encoder and inference pipeline in visual tracking. Moreover, the effectiveness of discriminative trackers remains constrained due to the adoption of the dual-branch pipeline. To tackle the inferior effectiveness of vanilla ViT, we propose an Adaptive ViT Model Prediction tracker (AViTMP) to design a customised tracking method. This method bridges the single-branch network with discriminative models for the first time. Specifically, in the proposed encoder AViT encoder, we introduce a tracking-tailored Adaptor module for vanilla ViT and a joint target state embedding to enrich the target-prior embedding paradigm. Then, we combine the AViT encoder with a discriminative transformer-specific model predictor to predict the accurate location. Furthermore, to mitigate the limitations of conventional inference practice, we present a novel inference pipeline called CycleTrack, which bolsters the tracking robustness in the presence of distractors via bidirectional cycle tracking verification. In the experiments, we evaluated AViTMP on eight tracking benchmarks for a comprehensive assessment, including LaSOT, LaSOTExtSub, AVisT, etc. The experimental results unequivocally establish that, under fair comparison, AViTMP achieves state-of-the-art performance, especially in terms of long-term tracking and robustness. The source code will be released at https://github.com/Tchuanm/AViTMP.

cs.CV↗

arXiv2024

From Halos to Galaxies. IX. Estimate of Halo Assembly History for SDSS Galaxy Groups

The properties of the galaxies are tightly connected to their host halo mass and halo assembly history. Accurate measurement of the halo assembly history in observation is challenging but crucial to the understanding of galaxy formation and evolution. The stellar-to-halo mass ratio ($M_*/M_{\mathrm{h}}$) for the centrals has often been used to indicate the halo assembly time $t_{\mathrm{h,50}}$ of the group, where $t_{\mathrm{h,50}}$ is the lookback time at which a halo has assembled half of its present-day virial mass. Using mock data from the semi-analytic models, we find that $M_*/M_{\mathrm{h}}$ shows a significant scatter with $t_{\mathrm{h,50}}$, with a strong systematic difference between the group with a star-forming central (blue group) and passive central (red group). To improve the accuracy, we develop machine-learning models to estimate $t_{\mathrm{h,50}}$ for galaxy groups using only observable quantities in the mocks. Since star-formation quenching will decouple the co-growth of the dark matter and baryon, we train our models separately for blue and red groups. Our models have successfully recovered $t_{\mathrm{h,50}}$, within an accuracy of $\sim$ 1.09 Gyr. With careful calibrations of individual observable quantities in the mocks with SDSS observations, we apply the trained models to the SDSS Yang et al. groups and derive the $t_{\mathrm{h,50}}$ for each group for the first time. The derived SDSS $t_{\mathrm{h,50}}$ distributions are in good agreement with that in the mocks, in particular for blue groups. The derived halo assembly history, together with the halo mass, make an important step forward in studying the halo-galaxy connections in observation.

astro-ph.GA↗

arXiv2024

FAITH: Frequency-domain Attention In Two Horizons for Time Series Forecasting

Time Series Forecasting plays a crucial role in various fields such as industrial equipment maintenance, meteorology, energy consumption, traffic flow and financial investment. However, despite their considerable advantages over traditional statistical approaches, current deep learning-based predictive models often exhibit a significant deviation between their forecasting outcomes and the ground truth. This discrepancy is largely due to an insufficient emphasis on extracting the sequence's latent information, particularly its global information within the frequency domain and the relationship between different variables. To address this issue, we propose a novel model Frequency-domain Attention In Two Horizons, which decomposes time series into trend and seasonal components using a multi-scale sequence adaptive decomposition and fusion architecture, and processes them separately. FAITH utilizes Frequency Channel feature Extraction Module and Frequency Temporal feature Extraction Module to capture inter-channel relationships and temporal global information in the sequence, significantly improving its ability to handle long-term dependencies and complex patterns. Furthermore, FAITH achieves theoretically linear complexity by modifying the time-frequency domain transformation method, effectively reducing computational costs. Extensive experiments on 6 benchmarks for long-term forecasting and 3 benchmarks for short-term forecasting demonstrate that FAITH outperforms existing models in many fields, such as electricity, weather and traffic, proving its effectiveness and superiority both in long-term and short-term time series forecasting tasks. Our codes and data are available at https://github.com/LRQ577/FAITH.

cs.LG↗

arXiv2024

Exploring the Reversal Curse and Other Deductive Logical Reasoning in BERT and GPT-Based Large Language Models

The term "Reversal Curse" refers to the scenario where auto-regressive decoder large language models (LLMs), such as ChatGPT, trained on "A is B" fail to learn "B is A," assuming that B and A are distinct and can be uniquely identified from each other, demonstrating a basic failure of logical deduction. This raises a red flag in the use of GPT models for certain general tasks such as constructing knowledge graphs, considering their adherence to this symmetric principle. In our study, we examined a bidirectional LLM, BERT, and found that it is immune to the reversal curse. Driven by ongoing efforts to construct biomedical knowledge graphs with LLMs, we also embarked on evaluating more complex but essential deductive reasoning capabilities. This process included first training encoder and decoder language models to master the intersection and union operations on two sets and then moving on to assess their capability to infer different combinations of union and intersection operations on three newly created sets. The findings showed that while both encoder and decoder language models, trained for tasks involving two sets (union/intersection), were proficient in such scenarios, they encountered difficulties when dealing with operations that included three sets (various combinations of union and intersection). Our research highlights the distinct characteristics of encoder and decoder models in simple and complex logical reasoning. In practice, the choice between BERT and GPT should be guided by the specific requirements and nature of the task at hand, leveraging their respective strengths in bidirectional context comprehension and sequence prediction.

cs.CL↗

arXiv2024

CUPID: Improving Battle Fairness and Position Satisfaction in Online MOBA Games with a Re-matchmaking System

The multiplayer online battle arena (MOBA) genre has gained significant popularity and economic success, attracting considerable research interest within the Human-Computer Interaction community. Enhancing the gaming experience requires a deep understanding of player behavior, and a crucial aspect of MOBA games is matchmaking, which aims to assemble teams of comparable skill levels. However, existing matchmaking systems often neglect important factors such as players' position preferences and team assignment, resulting in imbalanced matches and reduced player satisfaction. To address these limitations, this paper proposes a novel framework called CUPID, which introduces a novel process called ``re-matchmaking'' to optimize team and position assignments to improve both fairness and player satisfaction. CUPID incorporates a pre-filtering step to ensure a minimum level of matchmaking quality, followed by a pre-match win-rate prediction model that evaluates the fairness of potential assignments. By simultaneously considering players' position satisfaction and game fairness, CUPID aims to provide an enhanced matchmaking experience. Extensive experiments were conducted on two large-scale, real-world MOBA datasets to validate the effectiveness of CUPID. The results surpass all existing state-of-the-art baselines, with an average relative improvement of 7.18% in terms of win prediction accuracy. Furthermore, CUPID has been successfully deployed in a popular online mobile MOBA game. The deployment resulted in significant improvements in match fairness and player satisfaction, as evidenced by critical Human-Computer Interaction (HCI) metrics covering usability, accessibility, and engagement, observed through A/B testing. To the best of our knowledge, CUPID is the first re-matchmaking system designed specifically for large-scale MOBA games.

cs.HC↗

arXiv2024

Methodology of Adapting Large English Language Models for Specific Cultural Contexts

The rapid growth of large language models(LLMs) has emerged as a prominent trend in the field of artificial intelligence. However, current state-of-the-art LLMs are predominantly based on English. They encounter limitations when directly applied to tasks in specific cultural domains, due to deficiencies in domain-specific knowledge and misunderstandings caused by differences in cultural values. To address this challenge, our paper proposes a rapid adaptation method for large models in specific cultural contexts, which leverages instruction-tuning based on specific cultural knowledge and safety values data. Taking Chinese as the specific cultural context and utilizing the LLaMA3-8B as the experimental English LLM, the evaluation results demonstrate that the adapted LLM significantly enhances its capabilities in domain-specific knowledge and adaptability to safety values, while maintaining its original expertise advantages.

cs.CL↗

arXiv2024

Design of ANF/MXene/SSG sandwich structure with electromagnetic shielding performance and impact resistance

Since entering the information era, electronic devices gradually play an important role in daily lives. However, the abuse of electronic devices leads to corresponding electromagnetic EM wave pollution. The complex external environment causes the potential for physical impact. In this work, an ANF MXene SSG flexible sandwich structure was fabricated according to methods of vacuum filtration, directional freeze-casting solidification, and polyurethane encapsulation. Apart from its excellent protection function, the sandwich structure also acts as a human body movement sensor.

physics.app-ph↗

arXiv2024

Extreme Limit Theory of Competing Risks under Power Normalization

Advanced science and technology provide a wealth of big data from different sources for extreme value analysis. Classical extreme value theory was extended to obtain an accelerated max-stable distribution family for modelling competing risk-based extreme data in Cao and Zhang (2021). In this paper, we establish probability models for power normalized maxima and minima from competing risks. The limit distributions consist of an extensional new accelerated max-stable and min-stable distribution family (termed as the accelerated p-max/p-min stable distribution), and its left-truncated version. The consistency and asymptotic normality are obtained for the maximum likelihood estimation of the parameters involved in the accelerated p-max and p-min stable distributions when it exists. The limit types of distributions are determined principally by the sample generating process and the interplay among the competing risks, which are illustrated by common examples. Further, the statistical inference concerning the maximum likelihood estimation and model diagnosis of this model was investigated. Numerical studies show first the efficient approximation of all limit scenarios as well as its comparable convergence rate in contrast with those under linear normalization, and then present the maximum likelihood estimation and diagnosis of accelerated p-max/p-min stable models for simulated data sets. Finally, two real datasets concerning annual maximum of ground level ozone and survival times of Stanford heart plant demonstrate the performance of our accelerated p-max and accelerated p-min stable models.

math.ST↗

arXiv2024

High-energy Neutrino Emission from NGC 1068 by Outflow-cloud Interactions

As the hottest high-energy neutrino spot, NGC 1068 has received much attention in recent years. Here we focus on the central region of the active galactic nuclei (AGN) and propose an outflow-cloud interaction model that could probably explain the observed neutrino data. Considering the accretion process adjacent to the central supermassive black hole (SMBH) of NGC 1068, strong outflows will be generated, which will likely interact with surrounding clouds floating in the corona region. Particles carried by the outflow will be accelerated to very high energy by the shocks forming during the outflow-cloud interactions. For the accelerated high-energy protons, $pγ$ interactions with the background photon field of the corona and disk and $pp$ interaction with the surrounding gas will produce considerable high-energy $γ$-rays and neutrino. However, because of the extremely dense photon fields in the corona and disk, the newly generated $γ$-rays will be significantly attenuated through the $γγ$ absorptions. In our scenario, the expected GeV-TeV $γ$-ray emission will be suppressed to a much lower level than the neutrino emission, consistent with the observational characteristics of NGC 1068, while the generated 1-30\,TeV neutrino flux can fit the IceCube data very well.

astro-ph.HE↗

arXiv2024

LLM as Prompter: Low-resource Inductive Reasoning on Arbitrary Knowledge Graphs

Knowledge Graph (KG) inductive reasoning, which aims to infer missing facts from new KGs that are not seen during training, has been widely adopted in various applications. One critical challenge of KG inductive reasoning is handling low-resource scenarios with scarcity in both textual and structural aspects. In this paper, we attempt to address this challenge with Large Language Models (LLMs). Particularly, we utilize the state-of-the-art LLMs to generate a graph-structural prompt to enhance the pre-trained Graph Neural Networks (GNNs), which brings us new methodological insights into the KG inductive reasoning methods, as well as high generalizability in practice. On the methodological side, we introduce a novel pretraining and prompting framework ProLINK, designed for low-resource inductive reasoning across arbitrary KGs without requiring additional training. On the practical side, we experimentally evaluate our approach on 36 low-resource KG datasets and find that ProLINK outperforms previous methods in three-shot, one-shot, and zero-shot reasoning tasks, exhibiting average performance improvements by 20%, 45%, and 147%, respectively. Furthermore, ProLINK demonstrates strong robustness for various LLM promptings as well as full-shot scenarios.

cs.AI↗

arXiv2024

Navigating Complexity: Toward Lossless Graph Condensation via Expanding Window Matching

Graph condensation aims to reduce the size of a large-scale graph dataset by synthesizing a compact counterpart without sacrificing the performance of Graph Neural Networks (GNNs) trained on it, which has shed light on reducing the computational cost for training GNNs. Nevertheless, existing methods often fall short of accurately replicating the original graph for certain datasets, thereby failing to achieve the objective of lossless condensation. To understand this phenomenon, we investigate the potential reasons and reveal that the previous state-of-the-art trajectory matching method provides biased and restricted supervision signals from the original graph when optimizing the condensed one. This significantly limits both the scale and efficacy of the condensed graph. In this paper, we make the first attempt toward \textit{lossless graph condensation} by bridging the previously neglected supervision signals. Specifically, we employ a curriculum learning strategy to train expert trajectories with more diverse supervision signals from the original graph, and then effectively transfer the information into the condensed graph with expanding window matching. Moreover, we design a loss function to further extract knowledge from the expert trajectories. Theoretical analysis justifies the design of our method and extensive experiments verify its superiority across different datasets. Code is released at https://github.com/NUS-HPC-AI-Lab/GEOM.

cs.LG↗

arXiv2024

What is the best model? Application-driven Evaluation for Large Language Models

General large language models enhanced with supervised fine-tuning and reinforcement learning from human feedback are increasingly popular in academia and industry as they generalize foundation models to various practical tasks in a prompt manner. To assist users in selecting the best model in practical application scenarios, i.e., choosing the model that meets the application requirements while minimizing cost, we introduce A-Eval, an application-driven LLMs evaluation benchmark for general large language models. First, we categorize evaluation tasks into five main categories and 27 sub-categories from a practical application perspective. Next, we construct a dataset comprising 678 question-and-answer pairs through a process of collecting, annotating, and reviewing. Then, we design an objective and effective evaluation method and evaluate a series of LLMs of different scales on A-Eval. Finally, we reveal interesting laws regarding model scale and task difficulty level and propose a feasible method for selecting the best model. Through A-Eval, we provide clear empirical and engineer guidance for selecting the best model, reducing barriers to selecting and using LLMs and promoting their application and development. Our benchmark is publicly available at https://github.com/UnicomAI/DataSet/tree/main/TestData/GeneralAbility.

cs.CL↗

arXiv2024

AV-DiT: Efficient Audio-Visual Diffusion Transformer for Joint Audio and Video Generation

Recent Diffusion Transformers (DiTs) have shown impressive capabilities in generating high-quality single-modality content, including images, videos, and audio. However, it is still under-explored whether the transformer-based diffuser can efficiently denoise the Gaussian noises towards superb multimodal content creation. To bridge this gap, we introduce AV-DiT, a novel and efficient audio-visual diffusion transformer designed to generate high-quality, realistic videos with both visual and audio tracks. To minimize model complexity and computational costs, AV-DiT utilizes a shared DiT backbone pre-trained on image-only data, with only lightweight, newly inserted adapters being trainable. This shared backbone facilitates both audio and video generation. Specifically, the video branch incorporates a trainable temporal attention layer into a frozen pre-trained DiT block for temporal consistency. Additionally, a small number of trainable parameters adapt the image-based DiT block for audio generation. An extra shared DiT block, equipped with lightweight parameters, facilitates feature interaction between audio and visual modalities, ensuring alignment. Extensive experiments on the AIST++ and Landscape datasets demonstrate that AV-DiT achieves state-of-the-art performance in joint audio-visual generation with significantly fewer tunable parameters. Furthermore, our results highlight that a single shared image generative backbone with modality-specific adaptations is sufficient for constructing a joint audio-video generator. Our source code and pre-trained models will be released.

cs.CV↗