arXiv subjects

Gang Liu

Publications and source records attributed to Gang Liu.

At least 109 records · Page 6Linked to original sources

arXiv2024

Detection of gamma-ray burst Amati relation based on Hubble data set and Pantheon+ samples

Using gamma-ray bursts as standard candles for cosmological parameter constraints rely on their empirical luminosity relations and low-redshift calibration. In this paper, we examine the Amati relation and its potential corrections based on the A118 sample of higher-quality gamma-ray bursts, using both Hubble data set and Pantheon+ samples as calibration samples in the redshift range of z < 1.965. In calibrating gamma-ray bursts using these two datasets, we employ Gaussian processes to obtain corresponding Hubble diagrams to avoid the dependence on cosmological models in the calibration process. We first divided the low-redshift sample of GRBs into two bins and examined the Amati relation and its potential modifications. We found that under both calibrations, the Amati relation did not show evidence of redshift evolution (68% confidence level). For the other two Amati relations that include redshift evolution terms, the central values of the redshift evolution coefficients deviated from 0, but due to the limitations of the sample size and the increase in the number of parameters, most of the redshift evolution coefficients were not able to be excluded from 0 at the 1 sigma level. Therefore, to assess their situation across the entire redshift range, we employed MCMC to globally fit three types of Amati relations. By computing AIC and BIC, we found that for the GRB A118 sample, the standard Amati relation remains the most fitting empirical luminosity formula, and no potential redshift evolution trend was observed for two different low-redshift calibrating sources.

astro-ph.CO↗

arXiv2024

Video-CCAM: Enhancing Video-Language Understanding with Causal Cross-Attention Masks for Short and Long Videos

Multi-modal large language models (MLLMs) have demonstrated considerable potential across various downstream tasks that require cross-domain knowledge. MLLMs capable of processing videos, known as Video-MLLMs, have attracted broad interest in video-language understanding. However, videos, especially long videos, contain more visual tokens than images, making them difficult for LLMs to process. Existing works either downsample visual features or extend the LLM context size, risking the loss of high-resolution information or slowing down inference speed. To address these limitations, we apply cross-attention layers in the intermediate projector between the visual encoder and the large language model (LLM). As the naive cross-attention mechanism is insensitive to temporal order, we further introduce causal cross-attention masks (CCAMs) within the cross-attention layers. This Video-MLLM, named Video-CCAM, is trained in a straightforward two-stage fashion: feature alignment and visual instruction tuning. We develop several Video-CCAM models based on LLMs of different sizes (4B, 9B, and 14B). Video-CCAM proves to be a robust Video-MLLM and shows outstanding performance from short videos to long ones. Among standard video benchmarks like MVBench and VideoChatGPT-QA, Video-CCAM shows outstanding performances (1st/2nd/3rd in MVBench and TGIF-QA, 2nd/3rd/4th in MSVD-QA, MSRVTT-QA, and ActivityNet-QA). In benchmarks encompassing long videos, Video-CCAM models can be directly adapted to long video understanding and still achieve exceptional scores despite being trained solely with images and 16-frame videos. Using 96 frames (6$\times$ the training number of frames), Video-CCAM models rank 1st/2nd/3rd in VideoVista and 1st/2nd/4th in MLVU among all open-source Video-MLLMs, respectively. The code is publicly available in \url{https://github.com/QQ-MM/Video-CCAM}.

cs.CV↗

arXiv2024

On Line-Separable Weighted Unit-Disk Coverage and Related Problems

Given a set $P$ of $n$ points and a set $S$ of $n$ weighted disks in the plane, the disk coverage problem is to compute a subset of disks of smallest total weight such that the union of the disks in the subset covers all points of $P$. The problem is NP-hard. In this paper, we consider a line-separable unit-disk version of the problem where all disks have the same radius and their centers are separated from the points of $P$ by a line $\ell$. We present an $O(n^{3/2}\log^2 n)$ time algorithm for the problem. This improves the previously best work of $O(n^2\log n)$ time. Our result leads to an algorithm of $O(n^{{7}/{2}}\log^2 n)$ time for the halfplane coverage problem (i.e., using $n$ weighted halfplanes to cover $n$ points), an improvement over the previous $O(n^4\log n)$ time solution. If all halfplanes are lower ones, our algorithm runs in $O(n^{{3}/{2}}\log^2 n)$ time, while the previous best algorithm takes $O(n^2\log n)$ time. Using duality, the hitting set problems under the same settings can be solved with similar time complexities.

cs.CG↗

arXiv2024

Unweighted Geometric Hitting Set for Line-Constrained Disks and Related Problems

Given a set $P$ of $n$ points and a set $S$ of $m$ disks in the plane, the disk hitting set problem asks for a smallest subset of $P$ such that every disk of $S$ contains at least one point in the subset. The problem is NP-hard. In this paper, we consider a line-constrained version in which all disks have their centers on a line. We present an $O(m\log^2n+(n+m)\log(n+m))$ time algorithm for the problem. This improves the previously best result of $O(m^2\log m+(n+m)\log(n+m))$ time for the weighted case of the problem where every point of $P$ has a weight and the objective is to minimize the total weight of the hitting set. Our algorithm actually solves a more general line-separable problem with a single intersection property: The points of $P$ and the disk centers are separated by a line $\ell$ and the boundary of every two disks intersect at most once on the side of $\ell$ containing $P$.

cs.CG↗

arXiv2024

Kinetically Coupled Scalar Fields Model and Cosmological Tensions

In this paper, we investigate the kinetically coupled early dark energy (EDE) and scalar field dark matter to address cosmological tensions. The EDE model presents an intriguing theoretical approach to resolving the Hubble tension, but it exacerbates the large-scale structure tension. We consider the interaction between dark matter and EDE, such that the drag of dark energy on dark matter suppresses structure growth, which can alleviate large-scale structure tension. We replace cold dark matter with scalar field dark matter, which has the property of suppressing structure growth on small scales. We employed the Markov Chain Monte Carlo method to constrain the model parameters, our new model reveals a non-zero coupling constant of $0.030 \pm 0.026$ at a 68% confidence level. The coupled model yields the Hubble constant value of $72.38^{+0.71}_{-0.82}$ km/s/Mpc, which resolves the Hubble tension. However, similar to the EDE model, it also obtains a larger $S_8$ value compared to the $Λ$CDM model, further exacerbating the large-scale structure tension. The EDE model and the new model yield the best-fit values of $0.8316$ and $0.8146$ for $S_8$, respectively, indicating that the new model partially alleviates the negative effect of the EDE model. However, this signature disappears when comparing marginalised posterior probabilities, and both models produce similar results. The values obtained from the EDE model and the new model are $0.822^{+0.011}_{-0.0093}$ and $0.819^{+0.013}_{-0.0092}$, respectively, at a 68% confidence level.

astro-ph.CO↗

arXiv2024

Cosmological Constraints on Thermal Friction of Axion Dark Matter

In this paper, we investigate the process in which axion dark matter undergoes thermal friction, resulting in energy injection into dark radiation, with the aim of mitigating the Hubble tension and large-scale structure tension. In the early universe, this scenario led to a rapid increase in the energy density of dark radiation; in the late universe, the evolution of axion dark matter is similar to that of cold dark matter, with this scenario resembling decaying dark matter and serving to ease the large-scale structure tension. We employ cosmological observational data, including cosmic microwave background (CMB), baryon acoustic oscillation (BAO), supernova data (SNIa), $H_0$ measurement from SH0ES, and $S_8$ from the Dark Energy Survey Year-3 (DES), to study and analyze this model. Our results indicate that the thermal friction model offers partial alleviation of the large-scale structure tension, while its contribution on alleviating Hubble tension can be ignored. The new model yields the value of $S_8$ is $0.795\pm 0.011$ at a 68% confidence level, while the $Λ$CDM model yields a result of $0.8023\pm 0.0085$. In addition, the new model exhibits a lower $χ^2_\mathrm{tot}$ value, with a difference of -2.60 compared to the $Λ$CDM model. Additionally, we incorporate Lyman-$α$ data to re-constrain the new model and find a slight improvement in the results, with the values of $H_0$ and $S_8$ being $68.76^{+0.39}_{-0.35}$ km/s/Mpc and $0.791\pm 0.011$ at a 68% confidence level, respectively.

astro-ph.CO↗

arXiv2024

Mitigating Cosmological Tensions via Momentum-Coupled Dark Sector Model

In this paper, we investigate the momentum coupling between early dark energy (EDE) and cold dark matter to alleviate cosmological tensions. EDE has exhibited promising efficacy in addressing the Hubble tension, but it exacerbates the large-scale structure tension. We consider the interaction between EDE and cold dark matter, introducing a pure momentum exchange between them to alleviate the large-scale structure tension introduced by the EDE model. We find that this coupling model is consistent with the EDE model, yielding a higher value for $H_0$, which can resolve the Hubble tension. Additionally, the new model exhibits a suppressive effect on structure growth, contributing to the alleviation of the large-scale structure tension. By utilizing the Markov Chain Monte Carlo method and incorporating various cosmological data, the coupling model constrains the best-fit values for $H_0$ to be $72.23$ km/s/Mpc and for $S_8$ to be 0.8192. Compared to the $Λ$CDM model, the new models have not fully resolved the large-scale structure tension. However, in contrast to the best-fit value of 0.8316 for $S_8$ obtained from the EDE model, the new model alleviates the negative impact of the EDE model.

astro-ph.CO↗

arXiv2024

Coupled Dark Sector Models and Cosmological Tensions

In this paper, we introduce two coupling models of early dark energy (EDE) and cold dark matter aimed at alleviating cosmological tensions. We utilize the EDE component in the coupling models to relieve the Hubble tension, while leveraging the interaction between dark matter and dark energy to alleviate the large-scale structure tension. The interaction is implemented in the form of pure momentum coupling and Yukawa coupling. We employed various cosmological datasets, including cosmic microwave background radiation, baryon acoustic oscillations, Type Ia supernovae, the local distance-ladder data (SH0ES), and the Dark Energy Survey Year-3 data, to analyze our models. We first exclude SH0ES data from the entire dataset to constrain the parameters of novel models. We observe that the constraints on $H_0$ from two coupling models are slightly higher than that from the $Λ$CDM model, but they exhibit a significant inconsistency with the SH0ES data, consistent with prior research findings in the EDE model. Subsequently, we incorporate SH0ES data to re-constrain the parameters of various models, our findings reveal that both coupling models yield best-fit values for $H_0$ approximately around $72.23$ km/s/Mpc, effectively mitigating the Hubble tension. Similar to the EDE model, the coupling models yield the $S_8$ values that still surpasses the result of the $Λ$CDM model. Nevertheless, the best-fit values for $S_8$ obtained with the two new models are 0.8192 and 0.8177, respectively, which are lower than the 0.8316 achieved by the EDE model. Consequently, although our coupling models fail to fully resolve the large-scale structure tension, they partially mitigate the adverse effect of the original EDE model.

astro-ph.CO↗

arXiv2024

Superior Polymeric Gas Separation Membrane Designed by Explainable Graph Machine Learning

Gas separation using polymer membranes promises to dramatically drive down the energy, carbon, and water intensity of traditional thermally driven separation, but developing the membrane materials is challenging. Here, we demonstrate a novel graph machine learning (ML) strategy to guide the experimental discovery of synthesizable polymer membranes with performances simultaneously exceeding the empirical upper bounds in multiple industrially important gas separation tasks. Two predicted candidates are synthesized and experimentally validated to perform beyond the upper bounds for multiple gas pairs (O2/N2, H2/CH4, and H2/N2). Notably, the O2/N2 separation selectivity is 1.6-6.7 times higher than existing polymer membranes. The molecular origin of the high performance is revealed by combining the inherent interpretability of our ML model, experimental characterization, and molecule-level simulation. Our study presents a unique explainable ML-experiment combination to tackle challenging energy material design problems in general, and the discovered polymers are beneficial for industrial gas separation.

cond-mat.mtrl-sci↗

arXiv2024

Complete Kähler manifolds with nonnegative Ricci curvature

We consider complete Kähler manifolds with nonnegative Ricci curvature. The main results are: 1. When the manifold has nonnegative bisectional curvature, we show that $\lim\limits_{r\to\infty}\frac{r^{2}}{vol(B(p, r))}\int_{B(p, r)}S$ exists. In other words, it depends only on the manifold. This solves a question of Ni. Also, we establish estimates among volume growth ratio, integral of scalar curvature, and the degree of polynomial growth holomorphic functions. The new point is that the estimates are sharp for any prescribed volume growth rate. 2. We discover a strong rigidity for complete Ricci flat Kähler metrics. Let $M^n (n\geq 2)$ be a complete Kähler manifold with nonnegative Ricci curvature and Euclidean volume growth. Assume either the curvature has quadratic decay, or the Kähler metric is $dd^c$-exact with quadratic decay of scalar curvature. If one tangent cone at infinity is Ricci flat, then $M$ is Ricci flat. In particular, the tangent cone is unique. In other words, we can test Ricci flatness of the manifold by checking one single tangent cone. This seems unexpected, since apriori, there is no equation on $M$ and the Bishop-Gromov volume comparison is not sharp on Ricci flat (nonflat) manifolds. Such result is in sharp contrast to the Riemannian setting: Colding and Naber showed that tangent cones are quite flexible when $Ric\geq 0$ and $|Rm|r^2<C$. This reveals subtle differences between Riemannian case and Kähler case. The result contains lots of examples, such as all noncompact Ricci flat Kähler surfaces of Euclidean volume growth (hyper-Kähler ALE 4-manifolds classified by Kronheimer), higher dimensional examples of Tian-Yau type, as well as an example with irregular cross section. It also covers Ricci flat Kähler metrics of Euclidean volume growth on Stein manifolds with $b_2 = 0$(e.g., $\mathbb{C}^n$).

math.DG↗

arXiv2024

Enhancing Generalization in Medical Visual Question Answering Tasks via Gradient-Guided Model Perturbation

Leveraging pre-trained visual language models has become a widely adopted approach for improving performance in downstream visual question answering (VQA) applications. However, in the specialized field of medical VQA, the scarcity of available data poses a significant barrier to achieving reliable model generalization. Numerous methods have been proposed to enhance model generalization, addressing the issue from data-centric and model-centric perspectives. Data augmentation techniques are commonly employed to enrich the dataset, while various regularization approaches aim to prevent model overfitting, especially when training on limited data samples. In this paper, we introduce a method that incorporates gradient-guided parameter perturbations to the visual encoder of the multimodality model during both pre-training and fine-tuning phases, to improve model generalization for downstream medical VQA tasks. The small perturbation is adaptively generated by aligning with the direction of the moving average gradient in the optimization landscape, which is opposite to the directions of the optimizer's historical updates. It is subsequently injected into the model's visual encoder. The results show that, even with a significantly smaller pre-training image caption dataset, our approach achieves competitive outcomes on both VQA-RAD and SLAKE datasets.

cs.CV↗

arXiv2024

Graph Learning for Parameter Prediction of Quantum Approximate Optimization Algorithm

In recent years, quantum computing has emerged as a transformative force in the field of combinatorial optimization, offering novel approaches to tackling complex problems that have long challenged classical computational methods. Among these, the Quantum Approximate Optimization Algorithm (QAOA) stands out for its potential to efficiently solve the Max-Cut problem, a quintessential example of combinatorial optimization. However, practical application faces challenges due to current limitations on quantum computational resource. Our work optimizes QAOA initialization, using Graph Neural Networks (GNN) as a warm-start technique. This sacrifices affordable computational resource on classical computer to reduce quantum computational resource overhead, enhancing QAOA's effectiveness. Experiments with various GNN architectures demonstrate the adaptability and stability of our framework, highlighting the synergy between quantum algorithms and machine learning. Our findings show GNN's potential in improving QAOA performance, opening new avenues for hybrid quantum-classical approaches in quantum computing and contributing to practical applications.

quant-ph↗

arXiv2024

SceMQA: A Scientific College Entrance Level Multimodal Question Answering Benchmark

The paper introduces SceMQA, a novel benchmark for scientific multimodal question answering at the college entrance level. It addresses a critical educational phase often overlooked in existing benchmarks, spanning high school to pre-college levels. SceMQA focuses on core science subjects including Mathematics, Physics, Chemistry, and Biology. It features a blend of multiple-choice and free-response formats, ensuring a comprehensive evaluation of AI models' abilities. Additionally, our benchmark provides specific knowledge points for each problem and detailed explanations for each answer. SceMQA also uniquely presents problems with identical contexts but varied questions to facilitate a more thorough and accurate assessment of reasoning capabilities. In the experiment, we evaluate both open-source and close-source state-of-the-art Multimodal Large Language Models (MLLMs), across various experimental settings. The results show that further research and development are needed in developing more capable MLLM, as highlighted by only 50% to 60% accuracy achieved by the strongest models. Our benchmark and analysis will be available at https://scemqa.github.io/

cs.AI↗

arXiv2024

On the Line-Separable Unit-Disk Coverage and Related Problems

Given a set $P$ of $n$ points and a set $S$ of $m$ disks in the plane, the disk coverage problem asks for a smallest subset of disks that together cover all points of $P$. The problem is NP-hard. In this paper, we consider a line-separable unit-disk version of the problem where all disks have the same radius and their centers are separated from the points of $P$ by a line $\ell$. We present an $O((n+m)\log(n+m))$ time algorithm for the problem. This improves the previously best result of $O(nm+ n\log n)$ time. Our techniques also solve the line-constrained version of the problem, where centers of all disks of $S$ are located on a line $\ell$ while points of $P$ can be anywhere in the plane. Our algorithm runs in $O((n+m)\log (m+ n)+m \log m\log n)$ time, which improves the previously best result of $O(nm\log(m+n))$ time. In addition, our results lead to an algorithm of $O(n^3\log n)$ time for a half-plane coverage problem (given $n$ half-planes and $n$ points, find a smallest subset of half-planes covering all points); this improves the previously best algorithm of $O(n^4\log n)$ time. Further, if all half-planes are lower ones, our algorithm runs in $O(n\log n)$ time while the previously best algorithm takes $O(n^2\log n)$ time.

cs.CG↗

arXiv2024

The 120Gbps VCSEL Array Based Optical Transmitter (ATx) Development for the High-Luminosity LHC (HL-LHC) Experiments

The integration of a Verticle Cavity Surface-Emitting Laser (VCSEL) array and a driving Application-Specific Integrated Circuit (ASIC) in a custom optical array transmitter module (ATx) for operation in the detector front-end is constructed, assembled and tested. The ATx provides 12 parallel channels with each channel operating at 10 Gbps. The optical transmitter eye diagram passes the eye mask and the bit-error rate (BER) less than 1E-12 transmission is achieved at 10 Gbps/ch. The overall insertion loss including the radiation induced attenuation is sufficiently low to meet the proposed link budget requirement.

physics.ins-det↗

arXiv2024

Optical Data Transmission ASICs for the High-Luminosity LHC (HL-LHC) Experiments

We present the design and test results of two optical data transmission ASICs for the High-Luminosity LHC (HL-LHC) experiments. These ASICs include a two-channel serializer (LOCs2) and a single-channel Vertical Cavity Surface Emitting Laser (VCSEL) driver (LOCld1V2). Both ASICs are fabricated in a commercial 0.25-um Silicon-on-Sapphire (SoS) CMOS technology and operate at a data rate up to 8 Gbps per channel. The power consumption of LOCs2 and LOCld1V2 are 1.25 W and 0.27 W at 8-Gbps data rate, respectively. LOCld1V2 has been verified meeting the radiation-tolerance requirements for HL-LHC experiments.

physics.ins-det↗

arXiv2023

Low-complexity Resource Allocation for Uplink RSMA in Future 6G Wireless Networks

Uplink rate-splitting multiple access (RSMA) requires optimization of decoding order and power allocation, while decoding order is a discrete variable, and it is very complex to find the optimal decoding order if the number of users is large enough. This letter proposes a low-complexity user pairing-based resource allocation algorithm with the objective of minimizing the maximum latency. Closed-form expressions for power and bandwidth allocation for a given latency are first derived. Then a bisection method is used to determine the minimum latency and optimal resource allocation. Finally, the proposed algorithm is compared with unpaired RSMA using an exhaustive method to obtain the optimal decoding order, unpaired RSMA using a suboptimal decoding order, paired non-orthogonal multiple access (NOMA) and unpaired NOMA. The results show that our proposed algorithm outperforms NOMA and achieves similar performance to unpaired RSMA. In addition, the complexity of the proposed algorithm is significantly reduced.

cs.IT↗

arXiv2023

Explaining Tree Model Decisions in Natural Language for Network Intrusion Detection

Network intrusion detection (NID) systems which leverage machine learning have been shown to have strong performance in practice when used to detect malicious network traffic. Decision trees in particular offer a strong balance between performance and simplicity, but require users of NID systems to have background knowledge in machine learning to interpret. In addition, they are unable to provide additional outside information as to why certain features may be important for classification. In this work, we explore the use of large language models (LLMs) to provide explanations and additional background knowledge for decision tree NID systems. Further, we introduce a new human evaluation framework for decision tree explanations, which leverages automatically generated quiz questions that measure human evaluators' understanding of decision tree inference. Finally, we show LLM generated decision tree explanations correlate highly with human ratings of readability, quality, and use of background knowledge while simultaneously providing better understanding of decision boundaries.

cs.CL↗