arXiv subjects

Jin Chen

Publications and source records attributed to Jin Chen.

At least 73 records · Page 4Linked to original sources

arXiv2025

Dual Test-time Training for Out-of-distribution Recommender System

Deep learning has been widely applied in recommender systems, which has achieved revolutionary progress recently. However, most existing learning-based methods assume that the user and item distributions remain unchanged between the training phase and the test phase. However, the distribution of user and item features can naturally shift in real-world scenarios, potentially resulting in a substantial decrease in recommendation performance. This phenomenon can be formulated as an Out-Of-Distribution (OOD) recommendation problem. To address this challenge, we propose a novel Dual Test-Time-Training framework for OOD Recommendation, termed DT3OR. In DT3OR, we incorporate a model adaptation mechanism during the test-time phase to carefully update the recommendation model, allowing the model to specially adapt to the shifting user and item features. To be specific, we propose a self-distillation task and a contrastive task to assist the model learning both the user's invariant interest preferences and the variant user/item characteristics during the test-time phase, thus facilitating a smooth adaptation to the shifting features. Furthermore, we provide theoretical analysis to support the rationale behind our dual test-time training framework. To the best of our knowledge, this paper is the first work to address OOD recommendation via a test-time-training strategy. We conduct experiments on three datasets with various backbones. Comprehensive experimental results have demonstrated the effectiveness of DT3OR compared to other state-of-the-art baselines.

cs.IR↗

arXiv2025

DeepSeek-V3 Technical Report

We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. To achieve efficient inference and cost-effective training, DeepSeek-V3 adopts Multi-head Latent Attention (MLA) and DeepSeekMoE architectures, which were thoroughly validated in DeepSeek-V2. Furthermore, DeepSeek-V3 pioneers an auxiliary-loss-free strategy for load balancing and sets a multi-token prediction training objective for stronger performance. We pre-train DeepSeek-V3 on 14.8 trillion diverse and high-quality tokens, followed by Supervised Fine-Tuning and Reinforcement Learning stages to fully harness its capabilities. Comprehensive evaluations reveal that DeepSeek-V3 outperforms other open-source models and achieves performance comparable to leading closed-source models. Despite its excellent performance, DeepSeek-V3 requires only 2.788M H800 GPU hours for its full training. In addition, its training process is remarkably stable. Throughout the entire training process, we did not experience any irrecoverable loss spikes or perform any rollbacks. The model checkpoints are available at https://github.com/deepseek-ai/DeepSeek-V3.

cs.CL↗

arXiv2025

Adaptive Sampled Softmax with Inverted Multi-Index: Methods, Theory and Applications

The softmax function is a cornerstone of multi-class classification, integral to a wide range of machine learning applications, from large-scale retrieval and ranking models to advanced large language models. However, its computational cost grows linearly with the number of classes, which becomes prohibitively expensive in scenarios with millions or even billions of classes. The sampled softmax, which relies on self-normalized importance sampling, has emerged as a powerful alternative, significantly reducing computational complexity. Yet, its estimator remains unbiased only when the sampling distribution matches the true softmax distribution. To improve both approximation accuracy and sampling efficiency, we propose the MIDX Sampler, a novel adaptive sampling strategy based on an inverted multi-index approach. Concretely, we decompose the softmax probability into several multinomial probabilities, each associated with a specific set of codewords and the last associated with the residual score of queries, thus reducing time complexity to the number of codewords instead of the number of classes. To further boost efficiency, we replace the query-specific residual probability with a simple uniform distribution, simplifying the computation while retaining high performance. Our method is backed by rigorous theoretical analysis, addressing key concerns such as sampling bias, gradient bias, convergence rates, and generalization error bounds. The results demonstrate that a smaller divergence from the ideal softmax distribution leads to faster convergence and improved generalization. Extensive experiments on large-scale language models, sequential recommenders, and extreme multi-class classification tasks confirm that the MIDX-Sampler delivers superior effectiveness and efficiency compared to existing approaches.

cs.LG↗

arXiv2024

Multiscale Latent Diffusion Model for Enhanced Feature Extraction from Medical Images

Various imaging modalities are used in patient diagnosis, each offering unique advantages and valuable insights into anatomy and pathology. Computed Tomography (CT) is crucial in diagnostics, providing high-resolution images for precise internal organ visualization. CT's ability to detect subtle tissue variations is vital for diagnosing diseases like lung cancer, enabling early detection and accurate tumor assessment. However, variations in CT scanner models and acquisition protocols introduce significant variability in the extracted radiomic features, even when imaging the same patient. This variability poses considerable challenges for downstream research and clinical analysis, which depend on consistent and reliable feature extraction. Current methods for medical image feature extraction, often based on supervised learning approaches, including GAN-based models, face limitations in generalizing across different imaging environments. In response to these challenges, we propose LTDiff++, a multiscale latent diffusion model designed to enhance feature extraction in medical imaging. The model addresses variability by standardizing non-uniform distributions in the latent space, improving feature consistency. LTDiff++ utilizes a UNet++ encoder-decoder architecture coupled with a conditional Denoising Diffusion Probabilistic Model (DDPM) at the latent bottleneck to achieve robust feature extraction and standardization. Extensive empirical evaluations on both patient and phantom CT datasets demonstrate significant improvements in image standardization, with higher Concordance Correlation Coefficients (CCC) across multiple radiomic feature categories. Through these advancements, LTDiff++ represents a promising solution for overcoming the inherent variability in medical imaging data, offering improved reliability and accuracy in feature extraction processes.

eess.IV↗

arXiv2024

Trusted Unified Feature-Neighborhood Dynamics for Multi-View Classification

Multi-view classification (MVC) faces inherent challenges due to domain gaps and inconsistencies across different views, often resulting in uncertainties during the fusion process. While Evidential Deep Learning (EDL) has been effective in addressing view uncertainty, existing methods predominantly rely on the Dempster-Shafer combination rule, which is sensitive to conflicting evidence and often neglects the critical role of neighborhood structures within multi-view data. To address these limitations, we propose a Trusted Unified Feature-NEighborhood Dynamics (TUNED) model for robust MVC. This method effectively integrates local and global feature-neighborhood (F-N) structures for robust decision-making. Specifically, we begin by extracting local F-N structures within each view. To further mitigate potential uncertainties and conflicts in multi-view fusion, we employ a selective Markov random field that adaptively manages cross-view neighborhood dependencies. Additionally, we employ a shared parameterized evidence extractor that learns global consensus conditioned on local F-N structures, thereby enhancing the global integration of multi-view features. Experiments on benchmark datasets show that our method improves accuracy and robustness over existing approaches, particularly in scenarios with high uncertainty and conflicting views. The code will be made available at https://github.com/JethroJames/TUNED.

cs.CV↗

arXiv2024

Taurus Database: How to be Fast, Available, and Frugal in the Cloud

Using cloud Database as a Service (DBaaS) offerings instead of on-premise deployments is increasingly common. Key advantages include improved availability and scalability at a lower cost than on-premise alternatives. In this paper, we describe the design of Taurus, a new multi-tenant cloud database system. Taurus separates the compute and storage layers in a similar manner to Amazon Aurora and Microsoft Socrates and provides similar benefits, such as read replica support, low network utilization, hardware sharing and scalability. However, the Taurus architecture has several unique advantages. Taurus offers novel replication and recovery algorithms providing better availability than existing approaches using the same or fewer replicas. Also, Taurus is highly optimized for performance, using no more than one network hop on critical paths and exclusively using append-only storage, delivering faster writes, reduced device wear, and constant-time snapshots. This paper describes Taurus and provides a detailed description and analysis of the storage node architecture, which has not been previously available from the published literature.

cs.DB↗

arXiv2024

Enhancing Low Dose Computed Tomography Images Using Consistency Training Techniques

Diffusion models have significant impact on wide range of generative tasks, especially on image inpainting and restoration. Although the improvements on aiming for decreasing number of function evaluations (NFE), the iterative results are still computationally expensive. Consistency models are as a new family of generative models, enable single-step sampling of high quality data without the need for adversarial training. In this paper, we introduce the beta noise distribution, which provides flexibility in adjusting noise levels. This is combined with a sinusoidal curriculum that enhances the learning of the trajectory between the noise distribution and the posterior distribution of interest, allowing High Noise Improved Consistency Training (HN-iCT) to be trained in a supervised fashion. Additionally, High Noise Improved Consistency Training with Image Condition (HN-iCT-CN) architecture is introduced, enables to take Low Dose images as a condition for extracting significant features by Weighted Attention Gates (WAG).Our results indicate that unconditional image generation using HN-iCT significantly outperforms basic CT and iCT training techniques with NFE=1 on the CIFAR10 and CelebA datasets. Moreover, our image-conditioned model demonstrates exceptional performance in enhancing low-dose (LD) CT scans.

eess.IV↗

arXiv2024

Conformally invariant charge fluctuations in a strange metal

The strange metal is a peculiar phase of matter in which the electron scattering rate, $τ^{-1} \sim k_B T/\hbar$, which determines the electrical resistance, is universal across a wide family of materials and determined only by fundamental constants. In 1989, theorists hypothesized that this universality would manifest as scale-invariant behavior in the dynamic charge susceptibility, $χ''(q,ω)$. Here, we present momentum-resolved inelastic electron scattering measurements of the strange metal Bi$_2$Sr$_2$CaCu$_2$O$_{8+x}$ showing that the susceptibility has the scale-invariant form $χ''(q,ω) = T^{-ν} f(ω/T)$, with exponent $ν= 0.93$. We find the response is consistent with conformal invariance, meaning the dynamics may be thought of as occurring on a circle of radius $1/T$ in imaginary time, characterized by conformal dimension $Δ= 0.05$. Our study indicates that the strange metal is a universal phenomenon whose properties are not determined by microscopic properties of a particular material.

cond-mat.str-el↗

arXiv2024

Deep Group Interest Modeling of Full Lifelong User Behaviors for CTR Prediction

Extracting users' interests from their lifelong behavior sequence is crucial for predicting Click-Through Rate (CTR). Most current methods employ a two-stage process for efficiency: they first select historical behaviors related to the candidate item and then deduce the user's interest from this narrowed-down behavior sub-sequence. This two-stage paradigm, though effective, leads to information loss. Solely using users' lifelong click behaviors doesn't provide a complete picture of their interests, leading to suboptimal performance. In our research, we introduce the Deep Group Interest Network (DGIN), an end-to-end method to model the user's entire behavior history. This includes all post-registration actions, such as clicks, cart additions, purchases, and more, providing a nuanced user understanding. We start by grouping the full range of behaviors using a relevant key (like item_id) to enhance efficiency. This process reduces the behavior length significantly, from O(10^4) to O(10^2). To mitigate the potential loss of information due to grouping, we incorporate two categories of group attributes. Within each group, we calculate statistical information on various heterogeneous behaviors (like behavior counts) and employ self-attention mechanisms to highlight unique behavior characteristics (like behavior type). Based on this reorganized behavior data, the user's interests are derived using the Transformer technique. Additionally, we identify a subset of behaviors that share the same item_id with the candidate item from the lifelong behavior sequence. The insights from this subset reveal the user's decision-making process related to the candidate item, improving prediction accuracy. Our comprehensive evaluation, both on industrial and public datasets, validates DGIN's efficacy and efficiency.

cs.IR↗

arXiv2024

MSW-type compactifications of 6d $(1,0)$ SCFTs on 4-manifolds

In this work, we study compactifications of 6d $(1,0)$ SCFTs, in particular those of conformal matter type, on Kähler 4-manifolds. We show how this can be realized via wrapping M5 branes on 4-cycles of non-compact Calabi-Yau fourfolds with ADE singularity in the fiber. Such compactifications lead to domain walls in 3d $\mathcal{N}=2$ theories which flow to 2d $\mathcal{N}=(0,2)$ SCFTs. We compute the central charges of such 2d CFTs via 6d anomaly polynomials by employing a particular topological twist along the 4-manifold. Moreover, we study compactifications on non-compact 4-manifolds leading to coupled 3d-2d systems. We show how these can be glued together consistently to reproduce the central charge and anomaly polynomial obtained in the compact case. Lastly, we study concrete CFT proposals for some special cases.

hep-th↗

arXiv2024

Gyrokinetic Electromagnetic Particle Simulations in Triangular Meshes with C1 Finite Elements

The triangular mesh-based gyrokinetic scheme enables comprehensive axis-to-edge studies across the entire plasma volume. Our approach employs triangular finite elements with first-derivative continuity (C1), building on previous work to facilitate gyrokinetic simulations. Additionally, we have adopted the mixed variable/pullback scheme for gyrokinetic electromagnetic particle simulations. The filter-free treatment in the poloidal cross-section with triangular meshes introduces unique features and challenges compared to previous treatments using structured meshes. Our implementation has been validated through benchmarks using ITPA-TAE (Toroidicity-induced Alfvén Eigenmode) parameters, showing its capability in moderate to small electron skin depth regimes. Additional examinations using experimental parameters confirm its applicability to realistic plasma conditions.

physics.plasm-ph↗

arXiv2024

Improved algorithm for a two-dimensional Darwin particle-in-cell code

A two-dimensional particle-in-cell code for simulation of low-frequency electromagnetic processes in laboratory plasmas has been developed. The code uses the Darwin method omitting the electromagnetic wave propagation. The Darwin method separates the electric field into solenoidal and irrotational parts. The irrotational electric field is the electrostatic field calculated with the direct implicit algorithm. The solenoidal electric field is calculated with a new algorithm based on the equation for the electric field vorticity. The new algorithm is faster and more reliable than the Streamlined Darwin Field Formulation introduced decades ago. The system of linear equations in the new algorithm is solved using a standard iterative method. The code is applied to simulate an inductively coupled plasma with the driving current flowing around the plasma region.

physics.plasm-ph↗

arXiv2024

Modularity of Vafa-Witten Partition Functions from SymTFT

The 6d (2,0) theory of $N$ M5 branes compactified on the product geometry $T^2\times S$, where $S$ is a Kähler 4-manifold, can be studied in two different limits. In one limit, the size of $T^2$ is taken to zero and together with a topological twist one arrives at the Vafa-Witten partition function on $S$. On the other hand, taking the size of $S$ to zero leads to a 2d $\mathcal{N}=(0,4)$ theory. This gives rise to a 2d-4d correspondence where the Vafa-Witten partition functions are identified with the characters of the 2d theory. In this paper, we test this conjecture for Hirzebruch and Del Pezzo surfaces by employing the technique of SymTFT to show that the modular transformation properties of the two sides match. Moreover, we construct modular invariant 2d absolute partition functions and verify that they are invariant under gauging of a discrete symmetry at the self-dual point in coupling space. This provides further hints for the presence of duality defects in the 2d SCFT.

hep-th↗

arXiv2024

Efficient Transfer Learning Framework for Cross-Domain Click-Through Rate Prediction

Natural content and advertisement coexist in industrial recommendation systems but differ in data distribution. Concretely, traffic related to the advertisement is considerably sparser compared to that of natural content, which motivates the development of transferring knowledge from the richer source natural content domain to the sparser advertising domain. The challenges include the inefficiencies arising from the management of extensive source data and the problem of 'catastrophic forgetting' that results from the CTR model's daily updating. To this end, we propose a novel tri-level asynchronous framework, i.e., Efficient Transfer Learning Framework for Cross-Domain Click-Through Rate Prediction (E-CDCTR), to transfer comprehensive knowledge of natural content to advertisement CTR models. This framework consists of three key components: Tiny Pre-training Model ((TPM), which trains a tiny CTR model with several basic features on long-term natural data; Complete Pre-training Model (CPM), which trains a CTR model holding network structure and input features the same as target advertisement on short-term natural data; Advertisement CTR model (A-CTR), which derives its parameter initialization from CPM together with multiple historical embeddings from TPM as extra feature and then fine-tunes on advertisement data. TPM provides richer representations of user and item for both the CPM and A-CTR, effectively alleviating the forgetting problem inherent in the daily updates. CPM further enhances the advertisement model by providing knowledgeable initialization, thereby alleviating the data sparsity challenges typically encountered by advertising CTR models. Such a tri-level cross-domain transfer learning framework offers an efficient solution to address both data sparsity and `catastrophic forgetting', yielding remarkable improvements.

cs.IR↗

arXiv2024

Beyond Uncertainty: Evidential Deep Learning for Robust Video Temporal Grounding

Existing Video Temporal Grounding (VTG) models excel in accuracy but often overlook open-world challenges posed by open-vocabulary queries and untrimmed videos. This leads to unreliable predictions for noisy, corrupted, and out-of-distribution data. Adapting VTG models to dynamically estimate uncertainties based on user input can address this issue. To this end, we introduce SRAM, a robust network module that benefits from a two-stage cross-modal alignment task. More importantly, it integrates Deep Evidential Regression (DER) to explicitly and thoroughly quantify uncertainty during training, thus allowing the model to say "I do not know" in scenarios beyond its handling capacity. However, the direct application of traditional DER theory and its regularizer reveals structural flaws, leading to unintended constraints in VTG tasks. In response, we develop a simple yet effective Geom-regularizer that enhances the uncertainty learning framework from the ground up. To the best of our knowledge, this marks the first successful attempt of DER in VTG. Our extensive quantitative and qualitative results affirm the effectiveness, robustness, and interpretability of our modules and the uncertainty learning paradigm in VTG tasks. The code will be made available.

cs.CV↗

arXiv2024

Bulk-spatiotemporal vortex correspondence in gyromagnetic double-zero-index media

Photonic double-zero-index media, distinguished by concurrently zero-valued permittivity and permeability, exhibit extraordinary properties not found in nature. Remarkably, the notion of zero-index can be substantially expanded by generalizing the constitutive parameters from null scalars to nonreciprocal tensors with nonzero matrix elements but zero determinants. Here, we experimentally realize such a new class of gyromagnetic double-zero-index metamaterials possessing both double-zero-index features and nonreciprocal hallmarks. As an intrinsic property, this metamaterial always emerges at a spin-1/2 Dirac point of a topological phase transition. We discover and rigorously prove that a spatiotemporal reflection vortex singularity is always anchored to the metamaterial's Dirac point, with the vortex charge being determined by the topological invariant leap across the phase transition. This establishes a unique bulk-spatiotemporal vortex correspondence that extends the protected boundary effects into the time domain and exclusively characterizes topological phase transition points, setting it apart from any pre-existing bulk-boundary correspondence. Based on this correspondence, we propose and experimentally demonstrate a mechanism to deterministically generate optical spatiotemporal vortex pulses with firmly fixed central frequency and momentum, hence showing unparalleled robustness. Our findings uncover deep connections between zero-refractive-index photonics, topological photonics, and singular optics, opening the avenue for the manipulation of space-time topological light fields via the inherent topology of extreme-parameter metamaterials.

physics.optics↗

arXiv2024

DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model

We present DeepSeek-V2, a strong Mixture-of-Experts (MoE) language model characterized by economical training and efficient inference. It comprises 236B total parameters, of which 21B are activated for each token, and supports a context length of 128K tokens. DeepSeek-V2 adopts innovative architectures including Multi-head Latent Attention (MLA) and DeepSeekMoE. MLA guarantees efficient inference through significantly compressing the Key-Value (KV) cache into a latent vector, while DeepSeekMoE enables training strong models at an economical cost through sparse computation. Compared with DeepSeek 67B, DeepSeek-V2 achieves significantly stronger performance, and meanwhile saves 42.5% of training costs, reduces the KV cache by 93.3%, and boosts the maximum generation throughput to 5.76 times. We pretrain DeepSeek-V2 on a high-quality and multi-source corpus consisting of 8.1T tokens, and further perform Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) to fully unlock its potential. Evaluation results show that, even with only 21B activated parameters, DeepSeek-V2 and its chat versions still achieve top-tier performance among open-source models.

cs.CL↗

arXiv2024

Human-annotated label noise and their impact on ConvNets for remote sensing image scene classification

Convolutional neural networks (ConvNets) have been successfully applied to satellite image scene classification. Human-labeled training datasets are essential for ConvNets to perform accurate classification. Errors in human-annotated training datasets are unavoidable due to the complexity of satellite images. However, the distribution of real-world human-annotated label noises on remote sensing images and their impact on ConvNets have not been investigated. To fill this research gap, this study, for the first time, collected real-world labels from 32 participants and explored how their annotated label noise affect three representative ConvNets (VGG16, GoogleNet, and ResNet-50) for remote sensing image scene classification. We found that: (1) human-annotated label noise exhibits significant class and instance dependence; (2) an additional 1% of human-annotated label noise in training data leads to 0.5% reduction in the overall accuracy of ConvNets classification; (3) the error pattern of ConvNet predictions was strongly correlated with that of participant's labels. To uncover the mechanism underlying the impact of human labeling errors on ConvNets, we further compared it with three types of simulated label noise: uniform noise, class-dependent noise and instance-dependent noise. Our results show that the impact of human-annotated label noise on ConvNets significantly differs from all three types of simulated label noise, while both class dependence and instance dependence contribute to the impact of human-annotated label noise on ConvNets. These observations necessitate a reevaluation of the handling of noisy labels, and we anticipate that our real-world label noise dataset would facilitate the future development and assessment of label-noise learning algorithms.

cs.CV↗