arXiv subjects

Ping Chen

Publications and source records attributed to Ping Chen.

At least 91 records · Page 5Linked to original sources

arXiv2025

Multi-period Asset-liability Management with Reinforcement Learning in a Regime-Switching Market

This paper explores the mean-variance portfolio selection problem in a multi-period financial market characterized by regime-switching dynamics and uncontrollable liabilities. To address the uncertainty in the decision-making process within the financial market, we incorporate reinforcement learning (RL) techniques. Specifically, the study examines an exploratory mean-variance (EMV) framework where investors aim to minimize risk while maximizing returns under incomplete market information, influenced by shifting economic regimes. The market model includes risk-free and risky assets, with liability dynamics driven by a Markov regime-switching process. To align with real-world scenarios where financial decisions are made over discrete time periods, we adopt a multi-period dynamic model. We present an optimal portfolio strategy derived using RL techniques that adapt to these market conditions. The proposed solution addresses the inherent time inconsistency in classical mean-variance models by integrating a pre-committed strategy formulation. Furthermore, we incorporate partial market observability, employing stochastic filtering techniques to estimate unobservable market states. Numerical simulations and empirical tests on real financial data demonstrate that our method achieves superior returns, lower risk, and faster convergence compared to traditional models. These findings highlight the robustness and adaptability of our RL-based solution in dynamic and complex financial environments.

math.OC↗

arXiv2025

RevPRAG: Revealing Poisoning Attacks in Retrieval-Augmented Generation through LLM Activation Analysis

Retrieval-Augmented Generation (RAG) enriches the input to LLMs by retrieving information from the relevant knowledge database, enabling them to produce responses that are more accurate and contextually appropriate. It is worth noting that the knowledge database, being sourced from publicly available channels such as Wikipedia, inevitably introduces a new attack surface. RAG poisoning involves injecting malicious texts into the knowledge database, ultimately leading to the generation of the attacker's target response (also called poisoned response). However, there are currently limited methods available for detecting such poisoning attacks. We aim to bridge the gap in this work. Particularly, we introduce RevPRAG, a flexible and automated detection pipeline that leverages the activations of LLMs for poisoned response detection. Our investigation uncovers distinct patterns in LLMs' activations when generating correct responses versus poisoned responses. Our results on multiple benchmark datasets and RAG architectures show our approach could achieve 98% true positive rate, while maintaining false positive rates close to 1%.

cs.CR↗

arXiv2025

A search for minute-time-scale flares from the transient AT\,2024wpp

The AT 2018cow-like fast blue optical transient AT2022tsd showed a large number of few-minute-duration, high-luminosity (~10^43 erg/s) flares. We present an intensive search for such flares from another 18cow-like event, AT2024wpp. We have used the Large Array Survey Telescope (LAST) to observe this transient between 28 and 74 days after the approximate time of zero flux. The target was observed for about 23 hours to a sensitivity that allows one to detect 3x10^42 erg/s flares at S/N>5. No optical flares have been found, suggesting a one-sided 2-sigma confidence upper limit of <0.02 on the flare's duty cycle, and flare rate lower than about 0.11/hr. These limits suggest that not all 18cow-like objects display a high rate of minute-timescale luminous flares. This can be explained either by diversity in the 18cow-like population or by viewing angle effects (e.g., beaming), or rather that the optical depth towards the central emitting region did not fall below unity during the particular search window.

astro-ph.HE↗

arXiv2025

SafeCtrl: Region-Based Safety Control for Text-to-Image Diffusion via Detect-Then-Suppress

The widespread deployment of text-to-image models is challenged by their potential to generate harmful content. While existing safety methods, such as prompt rewriting or model fine-tuning, provide valuable interventions, they often introduce a trade-off between safety and fidelity. Recent localization-based approaches have shown promise, yet their reliance on explicit ``concept replacement" can sometimes lead to semantic incongruity. To address these limitations, we explore a more flexible detect-then-suppress paradigm. We introduce SafeCtrl, a lightweight, non-intrusive plugin that first precisely localizes unsafe content. Instead of performing a hard A-to-B substitution, SafeCtrl then suppresses the harmful semantics, allowing the generative process to naturally and coherently resolve into a safe, context-aware alternative. A key aspect of our work is a novel training strategy using Direct Preference Optimization (DPO). We leverage readily available, image-level preference data to train our module, enabling it to learn nuanced suppression behaviors and perform region-guided interventions at inference without requiring costly, pixel-level annotations. Extensive experiments show that SafeCtrl significantly outperforms state-of-the-art methods in both safety efficacy and fidelity preservation. Our findings suggest that decoupled, suppression-based control is a highly effective and scalable direction for building more responsible generative models.

cs.CV↗

arXiv2025

Adacc: An Adaptive Framework Unifying Compression and Activation Recomputation for LLM Training

Training large language models (LLMs) is often constrained by GPU memory limitations. To alleviate memory pressure, activation recomputation and data compression have been proposed as two major strategies. However, both approaches have limitations: recomputation introduces significant training overhead, while compression can lead to accuracy degradation and computational inefficiency when applied naively. In this paper, we propose Adacc, the first adaptive memory optimization framework that unifies activation recomputation and data compression to improve training efficiency for LLMs while preserving model accuracy. Unlike existing methods that apply static, rule-based strategies or rely solely on one technique, Adacc makes fine-grained, tensor-level decisions, dynamically selecting between recomputation, retention, and compression based on tensor characteristics and runtime hardware constraints. Adacc tackles three key challenges: (1) it introduces layer-specific compression algorithms that mitigate accuracy loss by accounting for outliers in LLM activations; (2) it employs a MILP-based scheduling policy to globally optimize memory strategies across layers; and (3) it integrates an adaptive policy evolution mechanism to update strategies during training in response to changing data distributions. Experimental results show that Adacc improves training throughput by 1.01x to 1.37x compared to state-of-the-art frameworks, while maintaining accuracy comparable to the baseline.

cs.LG↗

arXiv2025

Divide-Then-Rule: A Cluster-Driven Hierarchical Interpolator for Attribute-Missing Graphs

Deep graph clustering (DGC) for attribute-missing graphs is an unsupervised task aimed at partitioning nodes with incomplete attributes into distinct clusters. Addressing this challenging issue is vital for practical applications. However, research in this area remains underexplored. Existing imputation methods for attribute-missing graphs often fail to account for the varying amounts of information available across node neighborhoods, leading to unreliable results, especially for nodes with insufficient known neighborhood. To address this issue, we propose a novel method named Divide-Then-Rule Graph Completion (DTRGC). This method first addresses nodes with sufficient known neighborhood information and treats the imputed results as new knowledge to iteratively impute more challenging nodes, while leveraging clustering information to correct imputation errors. Specifically, Dynamic Cluster-Aware Feature Propagation (DCFP) initializes missing node attributes by adjusting propagation weights based on the clustering structure. Subsequently, Hierarchical Neighborhood-aware Imputation (HNAI) categorizes attribute-missing nodes into three groups based on the completeness of their neighborhood attributes. The imputation is performed hierarchically, prioritizing the groups with nodes that have the most available neighborhood information. The cluster structure is then used to refine the imputation and correct potential errors. Finally, Hop-wise Representation Enhancement (HRE) integrates information across multiple hops, thereby enriching the expressiveness of node representations. Experimental results on six widely used graph datasets show that DTRGC significantly improves the clustering performance of various DGC methods under attribute-missing graphs.

cs.LG↗

arXiv2025

EP250108a/SN 2025kg: A Jet-Driven Stellar Explosion Interacting With Circumstellar Material

We present optical, radio, and X-ray observations of EP250108a/SN 2025kg, a broad-line Type Ic supernova (SN Ic-BL) accompanying an Einstein Probe (EP) fast X-ray transient (FXT) at $z=0.176$. EP250108a/SN 2025kg possesses a double-peaked optical light curve and its spectrum transitions from a blue underlying continuum to a typical SN Ic-BL spectrum over time. We fit a radioactive decay model to the second peak of the optical light curve and find SN parameters that are consistent with the SNe Ic-BL population, while its X-ray and radio properties are consistent with those of low-luminosity GRB (LLGRB) 060218/SN 2006aj. We explore three scenarios to understand the system's multi-wavelength emission -- (a) SN ejecta interacting with an extended circumstellar medium (CSM), (b) the shocked cocoon of a collapsar-driven jet choked in its stellar envelope, and (c) the shocked cocoon of a collapsar-driven jet choked in an extended CSM. Models (b) and (c) can explain the optical light curve and are also consistent with the radio and X-ray observations. We favor model (c) because it can self-consistently explain both the X-ray prompt emission and first optical peak, but we do not rule out model (b). From the properties of the first peak in model (c), we find evidence that EP250108a/SN 2025kg interacts with an extended CSM, and infer an envelope mass $M_{\rm e} \sim 0.1\,\rm M_\odot$ and radius $R_{\rm e} \sim 4 \times 10^{13}$ cm. EP250108a/SN 2025kg's multi-wavelength properties make it a close analog to LLGRB 060218/SN 2006aj, and highlight the power of early follow-up observations in mapping the environments of massive stars prior to core collapse.

astro-ph.HE↗

arXiv2025

GeSn 320 \times 256 Focal Plane Array for Silicon-Based Short-wave Infrared Imaging

Short-wave infrared (SWIR) imaging arrays have demonstrated great potential in applications spanning from military to civilian consumer electronics. However, the current focal plane arrays (FPAs), which are based on compound semiconductors, have limited applications in civilian circumstances due to elevated manufacturing costs and prolonged fabrication cycle time. To address this, a high-performance 320 $\times$ 256 focal plane array based on group-IV semiconductors has been designed and manufactured on a Si substrate using a complementary metal-oxide semiconductor (CMOS) compatible fabrication process. The optical absorption layer is composed of GeSn alloy, whose bandgap could be tailored by choosing the appropriate Sn concentration. In this work, a 10% Sn concentration was employed, yielding a response cutoff wavelength of 2308 nm for the Si-based photodetector, which was measured at 298 K. Moreover, a specific detectivity of 9.7 $\times$ 10$^{11}$ cm$\cdot$ Hz$^{1/2}$ $\cdot$ W$^{-1}$ has been achieved at 77 K, surpassing all previously reported GeSn devices, and rivals commercial extended InGaAs photodetectors. With the help of read-out circuits (ROIC), SWIR images have been successfully captured for the first time by using Si-based GeSn FPA. This work demonstrates the potential of group IV imaging arrays for various applications in the commercial SWIR imaging field.

physics.optics↗

arXiv2025

RoSMM: A Robust and Secure Multi-Modal Watermarking Framework for Diffusion Models

Current image watermarking technologies are predominantly categorized into text watermarking techniques and image steganography; however, few methods can simultaneously handle text and image-based watermark data, which limits their applicability in complex digital environments. This paper introduces an innovative multi-modal watermarking approach, drawing on the concept of vector discretization in encoder-based vector quantization. By constructing adjacency matrices, the proposed method enables the transformation of text watermarks into robust image-based representations, providing a novel multi-modal watermarking paradigm for image generation applications. Additionally, this study presents a newly designed image restoration module to mitigate image degradation caused by transmission losses and various noise interferences, thereby ensuring the reliability and integrity of the watermark. Experimental results validate the robustness of the method under multiple noise attacks, providing a secure, scalable, and efficient solution for digital image copyright protection.

cs.MM↗

arXiv2025

Strong well-posedness of the two-dimensional stochastic Navier-Stokes equation on moving domains

In this paper, we establish the strong($H^1$) well-posedness of the two dimensional stochastic Navier-Stokes equation with multiplicative noise on moving domains. Due to the nonlocality effect, this equation exhibits a ``piecewise" variational setting. Namely the global well-posedness of this equation is decomposed into the well-posedness of a family of stochastic partial differential equations(SPDEs) in the variational setting on each small time-interval. We first examine the well-posedness on each time interval, which does not have (nonhomogeneous) coercivity. Subsequently, we give an estimate of lower bound of length of the time-interval, which enables us to achieve the global well-posedness.

math.PR↗

arXiv2025

A Dataset and Toolkit for Multiparameter Cardiovascular Physiology Sensing on Rings

Smart rings offer a convenient way to continuously and unobtrusively monitor cardiovascular physiological signals. However, a gap remains between the ring hardware and reliable methods for estimating cardiovascular parameters, partly due to the lack of publicly available datasets and standardized analysis tools. In this work, we present $τ$-Ring, the first open-source ring-based dataset designed for cardiovascular physiological sensing. The dataset comprises photoplethysmography signals (infrared and red channels) and 3-axis accelerometer data collected from two rings (reflective and transmissive optical paths), with 28.21 hours of raw data from 34 subjects across seven activities. $τ$-Ring encompasses both stationary and motion scenarios, as well as stimulus-evoked abnormal physiological states, annotated with four ground-truth labels: heart rate, respiratory rate, oxygen saturation, and blood pressure. Using our proposed RingTool toolkit, we evaluated three widely-used physics-based methods and four cutting-edge deep learning approaches. Our results show superior performance compared to commercial rings, achieving best MAE values of 5.18 BPM for heart rate, 2.98 BPM for respiratory rate, 3.22\% for oxygen saturation, and 13.33/7.56 mmHg for systolic/diastolic blood pressure estimation. The open-sourced dataset and toolkit aim to foster further research and community-driven advances in ring-based cardiovascular health sensing.

eess.IV↗

arXiv2025

Optimizing Large Model Training through Overlapped Activation Recomputation

Large model training often uses recomputation to alleviate memory pressure and pipelines to exploit the parallelism of data, tensors, and devices. However, existing recomputation approaches may incur high overhead when training real-world models, as they are executed on demand in the critical training path. In this paper, we present Lynx, a new recomputation framework to reduce overhead by overlapping recomputation with communication in training pipelines. To reduce the large search space for recomputation strategies, we propose a heuristic-based recomputation scheduling algorithm, which is based on the observation that there are identical structures in large DNN models so that we can apply the same scheduling policy to all such structures. Additionally, we propose a recomputation-aware model partitioning method to balance each stage's execution time for improved training throughput. Our comprehensive evaluation using GPT models with 1.3B-23B parameters shows that Lynx outperforms existing recomputation approaches by up to 1.37x.

cs.DC↗

arXiv2025

Mitigating Object Hallucinations in Large Vision-Language Models with Assembly of Global and Local Attention

Despite great success across various multimodal tasks, Large Vision-Language Models (LVLMs) often encounter object hallucinations with generated textual responses being inconsistent with the actual objects in images. We examine different LVLMs and pinpoint that one root cause of object hallucinations lies with deficient attention on discriminative image features. Specifically, LVLMs often predominantly attend to prompt-irrelevant global features instead of prompt-relevant local features, undermining their visual grounding capacity and leading to object hallucinations. We propose Assembly of Global and Local Attention (AGLA), a training-free and plug-and-play approach that mitigates hallucinations by assembling global features for response generation and local features for visual discrimination simultaneously. Specifically, we introduce an image-prompt matching scheme that captures prompt-relevant local features from images, leading to an augmented view of the input image where prompt-relevant content is highlighted while irrelevant distractions are suppressed. Hallucinations can thus be mitigated with a calibrated logit distribution that is from generative global features of the original image and discriminative local features of the augmented image. Extensive experiments show the superiority of AGLA in LVLM hallucination mitigation, demonstrating its wide applicability across both discriminative and generative tasks. Our code is available at https://github.com/Lackel/AGLA.

cs.CV↗

arXiv2025

Optimizing for the Shortest Path in Denoising Diffusion Model

In this research, we propose a novel denoising diffusion model based on shortest-path modeling that optimizes residual propagation to enhance both denoising efficiency and quality. Drawing on Denoising Diffusion Implicit Models (DDIM) and insights from graph theory, our model, termed the Shortest Path Diffusion Model (ShortDF), treats the denoising process as a shortest-path problem aimed at minimizing reconstruction error. By optimizing the initial residuals, we improve the efficiency of the reverse diffusion process and the quality of the generated samples. Extensive experiments on multiple standard benchmarks demonstrate that ShortDF significantly reduces diffusion time (or steps) while enhancing the visual fidelity of generated samples compared to prior arts. This work, we suppose, paves the way for interactive diffusion-based applications and establishes a foundation for rapid data generation. Code is available at https://github.com/UnicomAI/ShortDF.

cs.CV↗

arXiv2025

A Rule Based Solution to Co-reference Resolution in Clinical Text

Objective: The aim of this study was to build an effective co-reference resolution system tailored for the biomedical domain. Materials and Methods: Experiment materials used in this study is provided by the 2011 i2b2 Natural Language Processing Challenge. The 2011 i2b2 challenge involves coreference resolution in medical documents. Concept mentions have been annotated in clinical texts, and the mentions that co-refer in each document are to be linked by coreference chains. Normally, there are two ways of constructing a system to automatically discover co-referent links. One is to manually build rules for co-reference resolution, and the other category of approaches is to use machine learning systems to learn automatically from training datasets and then perform the resolution task on testing datasets. Results: Experiments show the existing co-reference resolution systems are able to find some of the co-referent links, and our rule based system performs well finding the majority of the co-referent links. Our system achieved 89.6% overall performance on multiple medical datasets. Conclusion: The experiment results show that manually crafted rules based on observation of training data is a valid way to accomplish high performance in this coreference resolution task for the critical biomedical domain.

cs.CL↗

arXiv2025

"Stones from Other Hills can Polish Jade": Zero-shot Anomaly Image Synthesis via Cross-domain Anomaly Injection

Industrial image anomaly detection (IAD) is a pivotal topic with huge value. Due to anomaly's nature, real anomalies in a specific modern industrial domain (i.e. domain-specific anomalies) are usually too rare to collect, which severely hinders IAD. Thus, zero-shot anomaly synthesis (ZSAS), which synthesizes pseudo anomaly images without any domain-specific anomaly, emerges as a vital technique for IAD. However, existing solutions are either unable to synthesize authentic pseudo anomalies, or require cumbersome training. Thus, we focus on ZSAS and propose a brand-new paradigm that can realize both authentic and training-free ZSAS. It is based on a chronically-ignored fact: Although domain-specific anomalies are rare, real anomalies from other domains (i.e. cross-domain anomalies) are actually abundant and directly applicable to ZSAS. Specifically, our new ZSAS paradigm makes three-fold contributions: First, we propose a novel method named Cross-domain Anomaly Injection (CAI), which directly exploits cross-domain anomalies to enable highly authentic ZSAS in a training-free manner. Second, to supply CAI with sufficient cross-domain anomalies, we build the first Domain-agnostic Anomaly Dataset within our best knowledge, which provides ZSAS with abundant real anomaly patterns. Third, we propose a CAI-guided Diffusion Mechanism, which further breaks the quantity limit of real anomalies and enable unlimited anomaly synthesis. Our head-to-head comparison with existing ZSAS solutions justifies our paradigm's superior performance for IAD and demonstrates it as an effective and pragmatic ZSAS solution.

cs.CV↗

arXiv2025

Toroidal graphs without $K_{5}^{-}$ and 6-cycles

Cai et al.\ proved that a toroidal graph $G$ without $6$-cycles is $5$-choosable, and proposed the conjecture that $\textsf{ch}(G) = 5$ if and only if $G$ contains a $K_{5}$ [J. Graph Theory 65 (2010) 1--15], where $\textsf{ch}(G)$ is the choice number of $G$. However, Choi later disproved this conjecture, and proved that toroidal graphs without $K_{5}^{-}$ (a $K_{5}$ missing one edge) and $6$-cycles are $4$-choosable [J. Graph Theory 85 (2017) 172--186]. In this paper, we provide a structural description, for toroidal graphs without $K_{5}^{-}$ and $6$-cycles. Using this structural description, we strengthen Choi's result in two ways: (I) we prove that such graphs have weak degeneracy at most three (nearly $3$-degenerate), and hence their DP-paint numbers and DP-chromatic numbers are at most four; (II) we prove that such graphs have Alon-Tarsi numbers at most $4$. Furthermore, all of our results are sharp in some sense.

math.CO↗

arXiv2025

SteLLA: A Structured Grading System Using LLMs with RAG

Large Language Models (LLMs) have shown strong general capabilities in many applications. However, how to make them reliable tools for some specific tasks such as automated short answer grading (ASAG) remains a challenge. We present SteLLA (Structured Grading System Using LLMs with RAG) in which a) Retrieval Augmented Generation (RAG) approach is used to empower LLMs specifically on the ASAG task by extracting structured information from the highly relevant and reliable external knowledge based on the instructor-provided reference answer and rubric, b) an LLM performs a structured and question-answering-based evaluation of student answers to provide analytical grades and feedback. A real-world dataset that contains students' answers in an exam was collected from a college-level Biology course. Experiments show that our proposed system can achieve substantial agreement with the human grader while providing break-down grades and feedback on all the knowledge points examined in the problem. A qualitative and error analysis of the feedback generated by GPT4 shows that GPT4 is good at capturing facts while may be prone to inferring too much implication from the given text in the grading task which provides insights into the usage of LLMs in the ASAG system.

cs.CL↗