arXiv subjects

Jiali Deng

Publications and source records attributed to Jiali Deng.

17 recordsLinked to original sources

arXiv2026

Pion structure in Holographic QCD

We employ a holographic model with a modified background that incorporates effective descriptions of key QCD features, including linear confinement and gluon condensation, to study the pion's internal structure, encompassing its mass spectrum as well as electromagnetic and gravitational form factors. This model is capable of simultaneously describing these diverse observables and reaches reasonable agreement with both experimental measurements and lattice QCD results. Our findings indicate that the model captures essential aspects of the pion. The description of multiple structure observables supports its potential as a useful tool for further investigations of pion properties.

hep-ph↗

arXiv2026

Proton Structure from a Soft-Wall Holographic QCD Model: Mass Spectrum, Form Factors, and Mechanical Properties

Understanding the internal structure of the proton-including its mass spectrum, electromagnetic and gravitational form factors, and mechanical properties-remains a central challenge in hadronic physics. While lattice QCD and experimental measurements provide valuable insights, a holographic framework with a single parameter set capable of simultaneously describing these diverse observables is still lacking. Here, we employ the soft wall model, a phenomenological holographic approach that incorporates gluon condensation and linear confinement, to compute the proton mass spectrum, electromagnetic form factors (EMFFs), and gravitational form factors (GFFs). Our results show good agreement with recent experimental data and lattice QCD calculations. Despite its phenomenological nature, the model's ability to simultaneously describe multiple observables suggests that it effectively mimics some key QCD features.

hep-ph↗

arXiv2026

Gravitational form factors of the pion in light-front holographic QCD

Understanding the internal structure of the pion-particularly the energy-momentum distributions of quarks and gluons and the internal mechanical properties encoded in its gravitational form factors-is a fundamental challenge in quantum chromodynamics (QCD). In this work, we study the gravitational form factors using light-front QCD (LFQCD), combined with the holographic QCD. Our main innovation is the introduction of an effective light-front wave function, with its five-dimensional component obtained from holographic QCD, which is then employed, within the light-front QCD framework, to calculate the pion's gravitational form factors $A(Q^2)$ and $D(Q^2)$ as well as its radius. Our computed pion gravitational form factors show good agreement with lattice QCD results, providing nontrivial support for the viability of our phenomenological model.

hep-ph↗

arXiv2026

Exploring Nucleon Structure and the Proton Mass Problem through Holographic QCD

Understanding the internal structure of the proton-including the distributions of quarks and gluons and their contributions to proton properties such as mass-remains a central challenge in quantum chromodynamics (QCD). While quark generalized parton distributions (GPDs) have been studied extensively, a unified approach that simultaneously extracts quark parton distribution functions (PDFs), gravitational form factors (GFFs), and gluon GPDs from experimental constraints is still lacking. Moreover, the role of gluons in proton mass generation, particularly through the trace anomaly mechanism, requires deeper theoretical and phenomenological exploration. In this study, we begin by extracting quark GPDs in protons using a parameterization method based on the electromagnetic form factors provided by Light-Front Holographic QCD (LFHQCD), from which we derive both quark PDFs and their GFFs. We then extend this approach to model gluon GPDs. Our calculations show consistency with experimental data and lattice QCD results and successfully reproduce soft Pomeron behavior. Furthermore, we investigate near-threshold $J/ψ$ production using gauge/string duality to quantify the contribution of the trace anomaly to the proton mass. Our results demonstrate that the parameterization method provides a consistent framework for describing both quark and gluon structure, bridging GPDs, PDFs, and GFFs. The analysis of $J/ψ$ production confirms that the trace anomaly contributes significantly ($\sim 24\%$) to the proton mass, with the calculated cross-section dependence on momentum transfer $t$ in agreement with experimental observations. This work advances the understanding of proton structure by integrating quark and gluon degrees of freedom and elucidating the origin of proton mass within QCD.

hep-ph↗

arXiv2026

Pricing mobility services under decision-dependent demand uncertainty: a carsharing case

The problem of pricing mobility services has attracted significant attention. In most studies, uncertain demand is modeled as an exogenous random variable with known distribution. This assumption overlooks the likely effect of prices on user adoption decisions. To address this dependency, we formulate the pricing problem as a stochastic program with decision-dependent demand uncertainty. Specifically, we make the non-standard assumption that the probability distribution of demand depends on pricing decisions. We show that the problem can be written as a mixed-integer linear program whose size is exponential in the input parameters. To find exact numerical solutions we specialize the L-shaped method for stochastic programs with decision-dependent uncertainty. In particular, we devise efficient separation routines by proving closed-form primal and dual solutions to the involved subproblems. In addition, we develop problem-specific valid inequalities and cut-sharing mechanisms which significantly improve convergence. We show that the method outperforms by far a commercial solver used to solve the monolithic formulation. Furthermore, in a case study based on a real-world carsharing system, we show that incorporating decision-dependent uncertainty improves expected profits by 8.39% compared to a benchmark that considers deterministic price-elastic demand, and by 8.53% compared to a benchmark that considers exogenous random demand, on average. In addition, we evaluate the performance of preventive pricing and relocation decisions under two vehicle allocation policies. The results suggest that a controlled allocation of vehicles to customers can improve service rates while only marginally affecting profits.

math.OC↗

arXiv2025

The nucleon structure from an AdS/QCD model in the Veneziano limit

We employ the VQCD model, a holographic approach that dynamically simulates essential QCD characteristics, including linear mass spectra, confinement, asymptotic freedom, and magnetic charge screening, while incorporating quark flavor effects. Using this model, we first calculate the proton mass spectrum and the wave function, incorporating anomalous dimensions to refine our results. Next, we compute the proton structure functions across a range of Bjorken $x$ values using consistent parameters. Furthermore, we derive the proton electromagnetic form factor by solving the electromagnetic field's motion equation, accounting for background effects, and demonstrate qualitative consistency with results from free electromagnetic fields coupled to fermions. Finally, we calculate the gravitational form factors by introducing an effective graviton mass $m$ arising from chiral symmetry breaking and the proton energy-momentum tensor. Our calculations yield results that are in excellent agreement with experimental data and lattice QCD computations, validating the VQCD model as a robust tool for studying proton properties.

nucl-th↗

arXiv2024

UniPSDA: Unsupervised Pseudo Semantic Data Augmentation for Zero-Shot Cross-Lingual Natural Language Understanding

Cross-lingual representation learning transfers knowledge from resource-rich data to resource-scarce ones to improve the semantic understanding abilities of different languages. However, previous works rely on shallow unsupervised data generated by token surface matching, regardless of the global context-aware semantics of the surrounding text tokens. In this paper, we propose an Unsupervised Pseudo Semantic Data Augmentation (UniPSDA) mechanism for cross-lingual natural language understanding to enrich the training data without human interventions. Specifically, to retrieve the tokens with similar meanings for the semantic data augmentation across different languages, we propose a sequential clustering process in 3 stages: within a single language, across multiple languages of a language family, and across languages from multiple language families. Meanwhile, considering the multi-lingual knowledge infusion with context-aware semantics while alleviating computation burden, we directly replace the key constituents of the sentences with the above-learned multi-lingual family knowledge, viewed as pseudo-semantic. The infusion process is further optimized via three de-biasing techniques without introducing any neural parameters. Extensive experiments demonstrate that our model consistently improves the performance on general zero-shot cross-lingual natural language understanding tasks, including sequence classification, information extraction, and question answering.

cs.CL↗

arXiv2023

Zonification and Pricing in Carsharing

In this article we address the problem of partitioning a carsharing business area into pricing zones. We formalize the problem mathematically and show that the resulting partitioning problem can be formulated as a binary integer programming problem. The partitioning problem is then extended to include pricing decisions, yielding the first joint zonification and pricing problem. The resulting mixed integer (possibly nonlinear) programming problem is solved exactly using an ad-hoc integer Benders decomposition for which we define effective problem-specific improvements. Extensive tests based on a real-world carsharing system demonstrate that the method outperforms a state-of-the-art commercial solver on problems of size comparable to those encountered in real-world instances. Furthermore, by jointly optimizing prices and pricing zones, we observe a profit increase of 7.01% compared to a zip code-based prior partition of the business area, and of 25.61% compared to a scenario where the business area is not partitioned.

math.OC↗

arXiv2022

Temporally Resolution Decrement: Utilizing the Shape Consistency for Higher Computational Efficiency

Image resolution that has close relations with accuracy and computational cost plays a pivotal role in network training. In this paper, we observe that the reduced image retains relatively complete shape semantics but loses extensive texture information. Inspired by the consistency of the shape semantics as well as the fragility of the texture information, we propose a novel training strategy named Temporally Resolution Decrement. Wherein, we randomly reduce the training images to a smaller resolution in the time domain. During the alternate training with the reduced images and the original images, the unstable texture information in the images results in a weaker correlation between the texture-related patterns and the correct label, naturally enforcing the model to rely more on shape properties that are robust and conform to the human decision rule. Surprisingly, our approach greatly improves both the training and inference efficiency of convolutional neural networks. On ImageNet classification, using only 33\% calculation quantity (randomly reducing the training image to 112$\times$112 within 90\% epochs) can still improve ResNet-50 from 76.32\% to 77.71\%. Superimposed with the strong training procedure of ResNet-50 on ImageNet, our method achieves 80.42\% top-1 accuracy with saving 37.5\% calculation overhead. To the best of our knowledge this is the highest ImageNet single-crop accuracy on ResNet-50 under 224$\times$224 without extra data or distillation.

cs.CV↗

arXiv2022

White Paper Assistance: A Step Forward Beyond the Shortcut Learning

The promising performances of CNNs often overshadow the need to examine whether they are doing in the way we are actually interested. We show through experiments that even over-parameterized models would still solve a dataset by recklessly leveraging spurious correlations, or so-called 'shortcuts'. To combat with this unintended propensity, we borrow the idea of printer test page and propose a novel approach called White Paper Assistance. Our proposed method involves the white paper to detect the extent to which the model has preference for certain characterized patterns and alleviates it by forcing the model to make a random guess on the white paper. We show the consistent accuracy improvements that are manifest in various architectures, datasets and combinations with other techniques. Experiments have also demonstrated the versatility of our approach on fine-grained recognition, imbalanced classification and robustness to corruptions.

cs.CV↗

arXiv2022

Selective Output Smoothing Regularization: Regularize Neural Networks by Softening Output Distributions

In this paper, we propose Selective Output Smoothing Regularization, a novel regularization method for training the Convolutional Neural Networks (CNNs). Inspired by the diverse effects on training from different samples, Selective Output Smoothing Regularization improves the performance by encouraging the model to produce equal logits on incorrect classes when dealing with samples that the model classifies correctly and over-confidently. This plug-and-play regularization method can be conveniently incorporated into almost any CNN-based project without extra hassle. Extensive experiments have shown that Selective Output Smoothing Regularization consistently achieves significant improvement in image classification benchmarks, such as CIFAR-100, Tiny ImageNet, ImageNet, and CUB-200-2011. Particularly, our method obtains 77.30% accuracy on ImageNet with ResNet-50, which gains 1.1% than baseline (76.2%). We also empirically demonstrate the ability of our method to make further improvements when combining with other widely used regularization techniques. On Pascal detection, using the SOSR-trained ImageNet classifier as the pretrained model leads to better detection performances.

cs.CV↗

arXiv2022

Channel Self-Supervision for Online Knowledge Distillation

Recently, researchers have shown an increased interest in the online knowledge distillation. Adopting an one-stage and end-to-end training fashion, online knowledge distillation uses aggregated intermediated predictions of multiple peer models for training. However, the absence of a powerful teacher model may result in the homogeneity problem between group peers, affecting the effectiveness of group distillation adversely. In this paper, we propose a novel online knowledge distillation method, \textbf{C}hannel \textbf{S}elf-\textbf{S}upervision for Online Knowledge Distillation (CSS), which structures diversity in terms of input, target, and network to alleviate the homogenization problem. Specifically, we construct a dual-network multi-branch structure and enhance inter-branch diversity through self-supervised learning, adopting the feature-level transformation and augmenting the corresponding labels. Meanwhile, the dual network structure has a larger space of independent parameters to resist the homogenization problem during distillation. Extensive quantitative experiments on CIFAR-100 illustrate that our method provides greater diversity than OKDDip and we also give pretty performance improvement, even over the state-of-the-art such as PCL. The results on three fine-grained datasets (StanfordDogs, StanfordCars, CUB-200-211) also show the significant generalization capability of our approach.

cs.CV↗

arXiv2022

Holographic deconfined QGP phase diagram and entropy with an anomalous flow in a magnetic field background

We assume that the initial hydrodynamic environment is a Quark Gluon Plasma (QGP) phase where merely u and d quarks are considered when adding a magnetic field. When considering the chiral magnetic effect in relativistic heavy ion collisions, an anomalous current will be formed in the QGP environment. The chiral magnetic current formed by these u and d quarks has an impact on the heavy quarkonium. By using fluid/gravity duality, the metric with anomalous flow is established by using fluid/gravity duality, so as to introduce the magnetic field effect into the corresponding metric. And then we use heavy quarkonium as a probe to study phase transition, and utilize the effective string tension of the heavy quarkonium to study phase transition by AdS/QCD theory. The characteristics of reporting inverse magnetic catalysis for the confinement-deconfinement transition with anomalous flow are in qualitative agreement with lattice QCD findings. The heavy-quarkonium asymptotic entropy distributions with different magnetic field around the confinement-deconfinement transition temperature are given in the paper.

hep-ph↗

arXiv2021

Cut-Thumbnail: A Novel Data Augmentation for Convolutional Neural Network

In this paper, we propose a novel data augmentation strategy named Cut-Thumbnail, that aims to improve the shape bias of the network. We reduce an image to a certain size and replace the random region of the original image with the reduced image. The generated image not only retains most of the original image information but also has global information in the reduced image. We call the reduced image as thumbnail. Furthermore, we find that the idea of thumbnail can be perfectly integrated with Mixed Sample Data Augmentation, so we put one image's thumbnail on another image while the ground truth labels are also mixed, making great achievements on various computer vision tasks. Extensive experiments show that Cut-Thumbnail works better than state-of-the-art augmentation strategies across classification, fine-grained image classification, and object detection. On ImageNet classification, ResNet-50 architecture with our method achieves 79.21\% accuracy, which is more than 2.8\% improvement on the baseline.

cs.CV↗

arXiv2021

Feature Mining: A Novel Training Strategy for Convolutional Neural Network

In this paper, we propose a novel training strategy for convolutional neural network(CNN) named Feature Mining, that aims to strengthen the network's learning of the local feature. Through experiments, we find that semantic contained in different parts of the feature is different, while the network will inevitably lose the local information during feedforward propagation. In order to enhance the learning of local feature, Feature Mining divides the complete feature into two complementary parts and reuse these divided feature to make the network learn more local information, we call the two steps as feature segmentation and feature reusing. Feature Mining is a parameter-free method and has plug-and-play nature, and can be applied to any CNN models. Extensive experiments demonstrate the wide applicability, versatility, and compatibility of our method.

cs.CV↗

arXiv2021

Go Small and Similar: A Simple Output Decay Brings Better Performance

Regularization and data augmentation methods have been widely used and become increasingly indispensable in deep learning training. Researchers who devote themselves to this have considered various possibilities. But so far, there has been little discussion about regularizing outputs of the model. This paper begins with empirical observations that better performances are significantly associated with output distributions, that have smaller average values and variances. By audaciously assuming there is causality involved, we propose a novel regularization term, called Output Decay, that enforces the model to assign smaller and similar output values on each class. Though being counter-intuitive, such a small modification result in a remarkable improvement on performance. Extensive experiments demonstrate the wide applicability, versatility, and compatibility of Output Decay.

cs.CV↗

arXiv2021

FocusedDropout for Convolutional Neural Network

In convolutional neural network (CNN), dropout cannot work well because dropped information is not entirely obscured in convolutional layers where features are correlated spatially. Except randomly discarding regions or channels, many approaches try to overcome this defect by dropping influential units. In this paper, we propose a non-random dropout method named FocusedDropout, aiming to make the network focus more on the target. In FocusedDropout, we use a simple but effective way to search for the target-related features, retain these features and discard others, which is contrary to the existing methods. We found that this novel method can improve network performance by making the network more target-focused. Besides, increasing the weight decay while using FocusedDropout can avoid the overfitting and increase accuracy. Experimental results show that even a slight cost, 10\% of batches employing FocusedDropout, can produce a nice performance boost over the baselines on multiple datasets of classification, including CIFAR10, CIFAR100, Tiny Imagenet, and has a good versatility for different CNN models.

cs.CV↗