arXiv subjects

Linh Nguyen

Publications and source records attributed to Linh Nguyen.

At least 19 recordsLinked to original sources

arXiv2026

CARB: A Characterization-Guided Framework for CNN Inference Cost Prediction and Deployment Screening

Accurate pre-deployment estimation of CNN inference cost--energy, latency, and peak memory--is increasingly critical as models are deployed on resource-constrained GPU platforms. Existing approaches rely on FLOPs, latency measurements, or single-device profiling as energy proxies, overlooking the non-linear interactions between architectural design and hardware load. We present a workload characterization study of 13 419 CNN configurations on two GPU platforms (RTX 5090 and RTX 3080) under GPU telemetry, revealing that energy, latency, and memory exhibit fundamentally distinct scaling behaviors: energy and latency diverge by 3x under high computational demand, and cross-GPU transferability differs by target--energy and latency require platform-specific models while memory transfers well across the two tested platforms. Building on these characterization findings, we develop CARB, a cascade-blended ensemble that jointly predicts all three targets with R2 ~0.99, and a two-stage deployment screening workflow that eliminates over 90% of candidates in seconds, reducing large design spaces to a Pareto-prioritized shortlist validated against real hardware.

cs.AR↗

arXiv2026

STRIKE: A Structured Taxonomy of Cybercrime for Risk, Impact, Knowledge, and Evolution

Cybercrime has grown exponentially in both scale and sophistication, posing significant threats. As attack methods evolve rapidly, traditional classification schemes often fail to capture the complexity and diversity of modern threats. To address this gap, we introduce STRIKE,a Structured Taxonomy for Risk, Impact, Knowledge, and Emerging Threats, which provides a unified, multi-dimensional framework for categorizing cybercrimes. STRIKE spans both conventional and emerging domains, including ransomware, phishing, network intrusion, child sexual abuse material (CSAM), cryptojacking, deepfakes, and supply chain attacks. It organizes threats using criteria such as attack vectors, adversarial tactics, societal impact, detection techniques, and mitigation strategies. Alongside the taxonomy, we review recent advances in detection methodologies and present a response workflow to assist practitioners under active threat conditions. This work offers researchers, security professionals, and policymakers a practical foundation for threat analysis, comparative evaluation, and adaptive cyber defense.

cs.CR↗

arXiv2026

ByteRover: Agent-Native Memory Through LLM-Curated Hierarchical Context

Memory-Augmented Generation (MAG) extends large language models with external memory to support long-context reasoning, but existing approaches universally treat memory as an external service that agents call into, delegating storage to separate pipelines of chunking, embedding, and graph extraction. This architectural separation means the system that stores knowledge does not understand it, leading to semantic drift between what the agent intended to remember and what the pipeline actually captured, loss of coordination context across agents, and fragile recovery after failures. In this paper, we propose ByteRover, an agent-native memory architecture that inverts the memory pipeline: the same LLM that reasons about a task also curates, structures, and retrieves knowledge. ByteRover represents knowledge in a hierarchical Context Tree, a file-based knowledge graph organized as Domain, Topic, Subtopic, and Entry, where each entry carries explicit relations, provenance, and an Adaptive Knowledge Lifecycle (AKL) with importance scoring, maturity tiers, and recency decay. Retrieval uses a 5-tier progressive strategy that resolves most queries at sub-100 ms latency without LLM calls, escalating to agentic reasoning only for novel questions. Experiments on LoCoMo and LongMemEval demonstrate that ByteRover achieves state-of-the-art accuracy on LoCoMo and competitive results on LongMemEval while requiring zero external infrastructure, no vector database, no graph database, no embedding service, with all knowledge stored as human-readable markdown files on the local filesystem.

cs.AI↗

arXiv2025

Quantum Algorithm for Estimating Ollivier-Ricci Curvature

We introduce a quantum algorithm for computing the Ollivier Ricci curvature, a discrete analogue of the Ricci curvature defined via optimal transport on graphs and general metric spaces. This curvature has seen applications ranging from signaling fragility in financial networks to serving as basic quantities in combinatorial quantum gravity. For inputs given as a point cloud with pairwise distances, we show that our algorithm can achieve an exponential speedup over the best-known classical methods for two particular classes of problem. Our work is another step toward quantum algorithms for geometrical problems that are capable of delivering practical value while also informing fundamental theory.

quant-ph↗

arXiv2025

SAND: A Self-supervised and Adaptive NAS-Driven Framework for Hardware Trojan Detection

The globalized semiconductor supply chain has made Hardware Trojans (HT) a significant security threat to embedded systems, necessitating the design of efficient and adaptable detection mechanisms. Despite promising machine learning-based HT detection techniques in the literature, they suffer from ad hoc feature selection and the lack of adaptivity, all of which hinder their effectiveness across diverse HT attacks. In this paper, we propose SAND, a selfsupervised and adaptive NAS-driven framework for efficient HT detection. Specifically, this paper makes three key contributions. (1) We leverage self-supervised learning (SSL) to enable automated feature extraction, eliminating the dependency on manually engineered features. (2) SAND integrates neural architecture search (NAS) to dynamically optimize the downstream classifier, allowing for seamless adaptation to unseen benchmarks with minimal fine-tuning. (3) Experimental results show that SAND achieves a significant improvement in detection accuracy (up to 18.3%) over state-of-the-art methods, exhibits high resilience against evasive Trojans, and demonstrates strong generalization.

cs.CR↗

arXiv2025

Exploring the Potential of Large Language Models in Fine-Grained Review Comment Classification

Code review is a crucial practice in software development. As code review nowadays is lightweight, various issues can be identified, and sometimes, they can be trivial. Research has investigated automated approaches to classify review comments to gauge the effectiveness of code reviews. However, previous studies have primarily relied on supervised machine learning, which requires extensive manual annotation to train the models effectively. To address this limitation, we explore the potential of using Large Language Models (LLMs) to classify code review comments. We assess the performance of LLMs to classify 17 categories of code review comments. Our results show that LLMs can classify code review comments, outperforming the state-of-the-art approach using a trained deep learning model. In particular, LLMs achieve better accuracy in classifying the five most useful categories, which the state-of-the-art approach struggles with due to low training examples. Rather than relying solely on a specific small training data distribution, our results show that LLMs provide balanced performance across high- and low-frequency categories. These results suggest that the LLMs could offer a scalable solution for code review analytics to improve the effectiveness of the code review process.

cs.SE↗

arXiv2025

Detecting Fraud in Financial Networks: A Semi-Supervised GNN Approach with Granger-Causal Explanations

Fraudulent activity in the financial industry costs billions annually. Detecting fraud, therefore, is an essential yet technically challenging task that requires carefully analyzing large volumes of data. While machine learning (ML) approaches seem like a viable solution, applying them successfully is not so easy due to two main challenges: (1) the sparsely labeled data, which makes the training of such approaches challenging (with inherent labeling costs), and (2) lack of explainability for the flagged items posed by the opacity of ML models, that is often required by business regulations. This article proposes SAGE-FIN, a semi-supervised graph neural network (GNN) based approach with Granger causal explanations for Financial Interaction Networks. SAGE-FIN learns to flag fraudulent items based on weakly labeled (or unlabelled) data points. To adhere to regulatory requirements, the flagged items are explained by highlighting related items in the network using Granger causality. We empirically validate the favorable performance of SAGE-FIN on a real-world dataset, Bipartite Edge-And-Node Attributed financial network (Elliptic++), with Granger-causal explanations for the identified fraudulent items without any prior assumption on the network structure.

q-fin.ST↗

arXiv2025

SuperRAG: Beyond RAG with Layout-Aware Graph Modeling

This paper introduces layout-aware graph modeling for multimodal RAG. Different from traditional RAG methods that mostly deal with flat text chunks, the proposed method takes into account the relationship of multimodalities by using a graph structure. To do that, a graph modeling structure is defined based on document layout parsing. The structure of an input document is retained with the connection of text chunks, tables, and figures. This representation allows the method to handle complex questions that require information from multimodalities. To confirm the efficiency of the graph modeling, a flexible RAG pipeline is developed using robust components. Experimental results on four benchmark test sets confirm the contribution of the layout-aware modeling for performance improvement of the RAG pipeline.

cs.CL↗

arXiv2024

Scalable AI Framework for Defect Detection in Metal Additive Manufacturing

Additive Manufacturing (AM) is transforming the manufacturing sector by enabling efficient production of intricately designed products and small-batch components. However, metal parts produced via AM can include flaws that cause inferior mechanical properties, including reduced fatigue response, yield strength, and fracture toughness. To address this issue, we leverage convolutional neural networks (CNN) to analyze thermal images of printed layers, automatically identifying anomalies that impact these properties. We also investigate various synthetic data generation techniques to address limited and imbalanced AM training data. Our models' defect detection capabilities were assessed using images of Nickel alloy 718 layers produced on a laser powder bed fusion AM machine and synthetic datasets with and without added noise. Our results show significant accuracy improvements with synthetic data, emphasizing the importance of expanding training sets for reliable defect detection. Specifically, Generative Adversarial Networks (GAN)-generated datasets streamlined data preparation by eliminating human intervention while maintaining high performance, thereby enhancing defect detection capabilities. Additionally, our denoising approach effectively improves image quality, ensuring reliable defect detection. Finally, our work integrates these models in the CLoud ADditive MAnufacturing (CLADMA) module, a user-friendly interface, to enhance their accessibility and practicality for AM applications. This integration supports broader adoption and practical implementation of advanced defect detection in AM processes.

cs.CV↗

arXiv2024

Provable Methods for Searching with an Imperfect Sensor

Assume that a target is known to be present at an unknown point among a finite set of locations in the plane. We search for it using a mobile robot that has imperfect sensing capabilities. It takes time for the robot to move between locations and search a location; we have a total time budget within which to conduct the search. We study the problem of computing a search path/strategy for the robot that maximizes the probability of detection of the target. Considering non-uniform travel times between points (e.g., based on the distance between them) is crucial for search and rescue applications; such problems have been investigated to a limited extent due to their inherent complexity. In this paper, we describe fast algorithms with performance guarantees for this search problem and some variants, complement them with complexity results, and perform experiments to observe their performance.

cs.RO↗

arXiv2024

Approximation Algorithms for Anchored Multiwatchman Routes

We study some variants of the $k$-\textsc{Watchman Routes} problem, the cooperative version of the classic \textsc{Watchman Routes} problem in a simple polygon. The watchmen may be required to see the whole polygon, or some pre-determined quota of area within the polygon, and we want to minimize the maximum length traveled by any watchman. While the single watchman version of the problem has received much attention is rather well understood, it is not the case for multiple watchmen version. We provide the first tight approximability results for the anchored $k$-\textsc{Watchman Routes} problem in a simple polygon, assuming $k$ is fixed, by a fully-polynomial time approximation scheme. The basis for the FPTAS is provided by an exact dynamic programming algorithm. If $k$ is a variable, we give constant-factor approximations.

cs.CG↗

arXiv2024

Coverage Path Planning For Minimizing Expected Time to Search For an Object With Continuous Sensing

In this paper, we present several results of both theoretical as well as practical interests. First, we propose the quota lawn mowing problem, an extension of the classic lawn mowing problem in computational geometry, as follows: given a quota of coverage, compute the shortest lawn mowing route to achieve said quota. We give constant-factor approximations for the quota lawn mowing problem. Second, we investigate the expected detection time minimization problem in geometric coverage path planning with local, continuous sensory information. We provide the first approximation algorithm with provable error bounds with pseudopolynomial running time. Our ideas also extend to another search mechanism, namely visibility-based search, which is related to the watchman route problem. We complement our theoretical analysis with some simple but effective heuristics for finding an object in minimum expected time, on which we provide simulation results.

cs.RO↗

arXiv2024

Revisiting the Disequilibrium Issues in Tackling Heart Disease Classification Tasks

In the field of heart disease classification, two primary obstacles arise. Firstly, existing Electrocardiogram (ECG) datasets consistently demonstrate imbalances and biases across various modalities. Secondly, these time-series data consist of diverse lead signals, causing Convolutional Neural Networks (CNNs) to become overfitting to the one with higher power, hence diminishing the performance of the Deep Learning (DL) process. In addition, when facing an imbalanced dataset, performance from such high-dimensional data may be susceptible to overfitting. Current efforts predominantly focus on enhancing DL models by designing novel architectures, despite these evident challenges, seemingly overlooking the core issues, therefore hindering advancements in heart disease classification. To address these obstacles, our proposed approach introduces two straightforward and direct methods to enhance the classification tasks. To address the high dimensionality issue, we employ a Channel-wise Magnitude Equalizer (CME) on signal-encoded images. This approach reduces redundancy in the feature data range, highlighting changes in the dataset. Simultaneously, to counteract data imbalance, we propose the Inverted Weight Logarithmic Loss (IWL) to alleviate imbalances among the data. When applying IWL loss, the accuracy of state-of-the-art models (SOTA) increases up to 5% in the CPSC2018 dataset. CME in combination with IWL also surpasses the classification results of other baseline models from 5% to 10%.

cs.LG↗

arXiv2024

Multirobot Watchman Routes in a Simple Polygon

The well-known \textsc{Watchman Route} problem seeks a shortest route in a polygonal domain from which every point of the domain can be seen. In this paper, we study the cooperative variant of the problem, namely the \textsc{$k$-Watchmen Routes} problem, in a simple polygon $P$. We look at both the version in which the $k$ watchmen must collectively see all of $P$, and the quota version in which they must see a predetermined fraction of $P$'s area. We give an exact pseudopolynomial time algorithm for the \textsc{$k$-Watchmen Routes} problem in a simple orthogonal polygon $P$ with the constraint that watchmen must move on axis-parallel segments, and there is a given common starting point on the boundary. Further, we give a fully polynomial-time approximation scheme and a constant-factor approximation for unconstrained movement. For the quota version, we give a constant-factor approximation in a simple polygon, utilizing the solution to the (single) \textsc{Quota Watchman Route} problem.

cs.CG↗

arXiv2024

Large language models can be zero-shot anomaly detectors for time series?

Recent studies have shown the ability of large language models to perform a variety of tasks, including time series forecasting. The flexible nature of these models allows them to be used for many applications. In this paper, we present a novel study of large language models used for the challenging task of time series anomaly detection. This problem entails two aspects novel for LLMs: the need for the model to identify part of the input sequence (or multiple parts) as anomalous; and the need for it to work with time series data rather than the traditional text input. We introduce sigllm, a framework for time series anomaly detection using large language models. Our framework includes a time-series-to-text conversion module, as well as end-to-end pipelines that prompt language models to perform time series anomaly detection. We investigate two paradigms for testing the abilities of large language models to perform the detection task. First, we present a prompt-based detection method that directly asks a language model to indicate which elements of the input are anomalies. Second, we leverage the forecasting capability of a large language model to guide the anomaly detection process. We evaluated our framework on 11 datasets spanning various sources and 10 pipelines. We show that the forecasting method significantly outperformed the prompting method in all 11 datasets with respect to the F1 score. Moreover, while large language models are capable of finding anomalies, state-of-the-art deep learning models are still superior in performance, achieving results 30% better than large language models.

cs.LG↗

arXiv2024

Spatially temporally distributed informative path planning for multi-robot systems

This paper investigates the problem of informative path planning for a mobile robotic sensor network in spatially temporally distributed mapping. The robots are able to gather noisy measurements from an area of interest during their movements to build a Gaussian Process (GP) model of a spatio-temporal field. The model is then utilized to predict the spatio-temporal phenomenon at different points of interest. To spatially and temporally navigate the group of robots so that they can optimally acquire maximal information gains while their connectivity is preserved, we propose a novel multistep prediction informative path planning optimization strategy employing our newly defined local cost functions. By using the dual decomposition method, it is feasible and practical to effectively solve the optimization problem in a distributed manner. The proposed method was validated through synthetic experiments utilizing real-world data sets.

cs.RO↗

arXiv2024

Optimizing Visibility-based Search in Polygonal Domains

Given a geometric domain $P$, visibility-based search problems seek routes for one or more mobile agents ("watchmen") to move within $P$ in order to be able to see a portion (or all) of $P$, while optimizing objectives, such as the length(s) of the route(s), the size (e.g., area or volume) of the portion seen, the probability of detecting a target distributed within $P$ according to a prior distribution, etc. The classic watchman route problem seeks a shortest route for an observer, with omnidirectional vision, to see all of $P$. In this paper we study bicriteria optimization problems for a single mobile agent within a polygonal domain $P$ in the plane, with the criteria of route length and area seen. Specifically, we address the problem of computing a minimum length route that sees at least a specified area of $P$ (minimum length, for a given area quota). We also study the problem of computing a length-constrained route that sees as much area as possible. We provide hardness results and approximation algorithms. In particular, for a simple polygon $P$ we provide the first fully polynomial-time approximation scheme for the problem of computing a shortest route seeing an area quota, as well as a (slightly more efficient) polynomial dual approximation. We also consider polygonal domains $P$ (with holes) and the special case of a planar domain consisting of a union of lines. Our results yield the first approximation algorithms for computing a time-optimal search route in $P$ to guarantee some specified probability of detection of a static target within $P$, randomly distributed in $P$ according to a given prior distribution.

cs.CG↗

arXiv2023

Intense squeezed light from lasers with sharply nonlinear gain at optical frequencies

Non-classical states of light, such as number-squeezed light, with fluctuations below the classical shot noise level, have important uses in metrology, communication, quantum information processing, and quantum simulation. However, generating these non-classical states of light, especially with high intensity and high degree of squeezing, is challenging. To address this problem, we introduce a new concept which uses gain to generate intense sub-Poissonian light at optical frequencies. It exploits a strongly nonlinear gain for photons which arises from a combination of frequency-dependent gain and Kerr nonlinearity. In this laser architecture, the interaction between the gain medium and Kerr nonlinearity suppresses the spontaneous emission at high photon number states, leading to a strong "negative feedback" that suppresses photon-number fluctuations. We discuss realistic implementations of this concept based on the use of solid-state gain media in laser cavities with Kerr nonlinear materials, showing how 90% squeezing of photon number fluctuations below the shot noise level can be realized.

quant-ph↗