arXiv subjects

Dongxin Liu

Publications and source records attributed to Dongxin Liu.

14 recordsLinked to original sources

arXiv2026

How Code Representation Shapes False-Positive Dynamics in Cross-Language LLM Vulnerability Detection

How code representation format shapes false positive behaviour in cross-language LLM vulnerability detection remains poorly understood. We systematically vary training intensity and code representation format, comparing raw source text with pruned Abstract Syntax Trees at both training time and inference time, across two 8B-parameter LLMs (Qwen3-8B and Llama 3.1-8B-Instruct) fine-tuned on C/C++ data from the NIST Juliet Test Suite (v1.3) and evaluated on Java (OWASP Benchmark v1.2) and Python (BenchmarkPython v0.1). Cross-language FPR reflects the joint effect of training-time and inference-time representation, not either alone. Text fine-tuning drives FPR upward monotonically (Qwen3-8B: 0.763 zero-shot, 0.866 pilot, 1.000 full-scale) while F1 remains stable (0.637-0.688), masking the collapse. We argue surface-cue memorisation is the primary mechanism: text fine-tuning encodes C/C++-specific API names and syntactic idioms as vulnerability triggers that fire indiscriminately on target-language code. A cross-representation probe, applying text-trained weights to AST-encoded input without retraining, isolates this: Qwen3-8B FPR drops from 0.866 to 0.583, and 37.2% of false positives revert to true negatives under AST input alone. Direct AST fine-tuning does not preserve the benefit (FPR at least 0.970), as flat linearisation introduces structural surface cues of its own. The pattern replicates across both model families. On BenchmarkPython the AST probe yields FPR=0.554, within 2.9 percentage points of the Java result, despite maximal surface-syntax differences, substantially weakening a domain-shift explanation. These findings motivate a pre-deployment consistency gate, running alerts through both text and AST paths, as a retraining-free filter for false-positive-sensitive settings, at the cost of reduced recall.

cs.CR↗

arXiv2024

Unveiling a crystal's entropy of disorder via electron diffraction. A statistical mechanics approach

Upon melting, the molecules in the crystal explore numerous configurations, reflecting an increase in disorder. The molar entropy of disorder can be defined by Bolzmann's formula dSd = Rln(Wd) where Wd is the increase in the number of microscopic states, so far inaccessible experimentally. We found that the Arrhenius frequency factor A of the electron diffraction signal decay provides Wd via an experimental equation A = AINTWd where AINT is an inelastic scattering cross-section. The method connects Clausius and Boltzmann experimentally and supplements the Clausius approach, being applicable to a femtogram quantity of thermally unstable and biomolecular crystals. The data also showed that crystal disordering and crystallization of melt are reciprocal, both governed by the entropy change, but manifesting in opposite directions.

physics.chem-ph↗

arXiv2023

FOCAL: Contrastive Learning for Multimodal Time-Series Sensing Signals in Factorized Orthogonal Latent Space

This paper proposes a novel contrastive learning framework, called FOCAL, for extracting comprehensive features from multimodal time-series sensing signals through self-supervised training. Existing multimodal contrastive frameworks mostly rely on the shared information between sensory modalities, but do not explicitly consider the exclusive modality information that could be critical to understanding the underlying sensing physics. Besides, contrastive frameworks for time series have not handled the temporal information locality appropriately. FOCAL solves these challenges by making the following contributions: First, given multimodal time series, it encodes each modality into a factorized latent space consisting of shared features and private features that are orthogonal to each other. The shared space emphasizes feature patterns consistent across sensory modalities through a modal-matching objective. In contrast, the private space extracts modality-exclusive information through a transformation-invariant objective. Second, we propose a temporal structural constraint for modality features, such that the average distance between temporally neighboring samples is no larger than that of temporally distant samples. Extensive evaluations are performed on four multimodal sensing datasets with two backbone encoders and two classifiers to demonstrate the superiority of FOCAL. It consistently outperforms the state-of-the-art baselines in downstream tasks with a clear margin, under different ratios of available labels. The code and self-collected dataset are available at https://github.com/tomoyoshki/focal.

cs.AI↗

arXiv2022

Hierarchical Overlapping Belief Estimation by Structured Matrix Factorization

Much work on social media opinion polarization focuses on a flat categorization of stances (or orthogonal beliefs) of different communities from media traces. We extend in this work in two important respects. First, we detect not only points of disagreement between communities, but also points of agreement. In other words, we estimate community beliefs in the presence of overlap. Second, in lieu of flat categorization, we consider hierarchical belief estimation, where communities might be hierarchically divided. For example, two opposing parties might disagree on core issues, but within a party, despite agreement on fundamentals, disagreement might occur on further details. We call the resulting combined problem a hierarchical overlapping belief estimation problem. To solve it, this paper develops a new class of unsupervised Non-negative Matrix Factorization (NMF) algorithms, we call Belief Structured Matrix Factorization (BSMF). Our proposed unsupervised algorithm captures both the latent belief intersections and dissimilarities, as well as a hierarchical structure. We discuss the properties of the algorithm and evaluate it on both synthetic and real-world datasets. In the synthetic dataset, our model reduces error by 40%. In real Twitter traces, it improves accuracy by around 10%. The model also achieves 96.08% self-consistency in a sanity check.

cs.SI↗

arXiv2022

Self-Contrastive Learning based Semi-Supervised Radio Modulation Classification

This paper presents a semi-supervised learning framework that is new in being designed for automatic modulation classification (AMC). By carefully utilizing unlabeled signal data with a self-supervised contrastive-learning pre-training step, our framework achieves higher performance given smaller amounts of labeled data, thereby largely reducing the labeling burden of deep learning. We evaluate the performance of our semi-supervised framework on a public dataset. The evaluation results demonstrate that our semi-supervised approach significantly outperforms supervised frameworks thereby substantially enhancing our ability to train deep neural networks for automatic modulation classification in a manner that leverages unlabeled data.

cs.LG↗

arXiv2022

Solving Nonsmooth Resource Allocation Problems with Feasibility Constraints through Novel Distributed Algorithms

The distributed non-smooth resource allocation problem over multi-agent networks is studied in this paper, where each agent is subject to globally coupled network resource constraints and local feasibility constraints described in terms of general convex sets. To solve such a problem, two classes of novel distributed continuous-time algorithms via differential inclusions and projection operators are proposed. Moreover, the convergence of the algorithms is analyzed by the Lyapunov functional theory and nonsmooth analysis. We illustrate that the first algorithm can globally converge to the exact optimum of the problem when the interaction digraph is weight-balanced and the local cost functions being strongly convex. Furthermore, the fully distributed implementation of the algorithm is studied over connected undirected graphs with strictly convex local cost functions. In addition, to improve the drawback of the first algorithm that requires initialization, we design the second algorithm which can be implemented without initialization to achieve global convergence to the optimal solution over connected undirected graphs with strongly convex cost functions. Finally, several numerical simulations verify the results.

math.OC↗

arXiv2022

Novel Distributed Algorithms Design for Nonsmooth Resource Allocation on Weight-Balanced Digraphs

In this paper, the distributed resource allocation problem on strongly connected and weight-balanced digraphs is investigated, where the decisions of each agent are restricted to satisfy the coupled network resource constraints and heterogeneous general convex sets. Moreover, the local cost function can be non-smooth. In order to achieve the exact optimum of the nonsmooth resource allocation problem, a novel continuous-time distributed algorithm based on the gradient descent scheme and differentiated projection operators is proposed. With the help of the set-valued LaSalle invariance principle and nonsmooth analysis, it is demonstrated that the algorithm converges asymptotically to the global optimal allocation. Moreover, for the situation where local constraints are not involved and the cost functions are differentiable with Lipschitz gradients, the convergence of the algorithm to the exact optimal solution is exponentially fast. Finally, the effectiveness of the proposed algorithms is illustrated by simulation examples.

math.OC↗

arXiv2022

Ionization and electron excitation of fullerene molecules in a carbon nanotube. A variable temperature/voltage transmission electron microscopic study

There is increasing attention to chemical applications of transmission electron microscopy, which is often plagued by radiation damage. The damage in organic matter predominantly occurs via ionization (radiolysis). Although radiolysis is highly important, previous studies on radiolysis have largely been descriptive and qualitative, lacking in such fundamental information as the product structure, the influence of the energy of the electrons, and the reaction kinetics. We need a chemically well-defined system to obtain such data, and have chosen as a model a variable-temperature and variable-voltage (VT/VV) study of the dimerization of a van der Waals dimer [60]fullerene (C60) to C120 in a carbon nanotube (CNT) as studied for individual reaction events at atomic resolution. We report here the identification of five reaction pathways that serve as mechanistic models of radiolysis damage. Two of them occur via a radical cation of the specimen generated by specimen ionization, and three involve singlet or triplet excited states of the specimen, as initiated by electron excitation of the CNT followed by energy transfer to the specimen. The pathways were distinguished by the pre-exponential factor and the Arrhenius activation energy. The prototypal reaction path is the radical cation reaction that we saw at <200 K, but, at >350 K, the excited-state reactions dominate. The results illustrate the importance of VT/VV kinetic analysis in the studies of radiation damage, and show that chemical ionization and electron excitation are inseparable but different mechanisms of radiation damage, which has so far been classified loosely under the single term "ionization."

physics.chem-ph↗

arXiv2021

DyDiff-VAE: A Dynamic Variational Framework for Information Diffusion Prediction

This paper describes a novel diffusion model, DyDiff-VAE, for information diffusion prediction on social media. Given the initial content and a sequence of forwarding users, DyDiff-VAE aims to estimate the propagation likelihood for other potential users and predict the corresponding user rankings. Inferring user interests from diffusion data lies the foundation of diffusion prediction, because users often forward the information in which they are interested or the information from those who share similar interests. Their interests also evolve over time as the result of the dynamic social influence from neighbors and the time-sensitive information gained inside/outside the social media. Existing works fail to model users' intrinsic interests from the diffusion data and assume user interests remain static along the time. DyDiff-VAE advances the state of the art in two directions: (i) We propose a dynamic encoder to infer the evolution of user interests from observed diffusion data. (ii) We propose a dual attentive decoder to estimate the propagation likelihood by integrating information from both the initial cascade content and the forwarding user sequence. Extensive experiments on four real-world datasets from Twitter and Youtube demonstrate the advantages of the proposed model; we show that it achieves 43.3% relative gains over the best baseline on average. Moreover, it has the lowest run-time compared with recurrent neural network based models.

cs.SI↗

arXiv2020

Scheduling Real-time Deep Learning Services as Imprecise Computations

The paper presents an efficient real-time scheduling algorithm for intelligent real-time edge services, defined as those that perform machine intelligence tasks, such as voice recognition, LIDAR processing, or machine vision, on behalf of local embedded devices that are themselves unable to support extensive computations. The work contributes to a recent direction in real-time computing that develops scheduling algorithms for machine intelligence tasks with anytime prediction. We show that deep neural network workflows can be cast as imprecise computations, each with a mandatory part and (several) optional parts whose execution utility depends on input data. The goal of the real-time scheduler is to maximize the average accuracy of deep neural network outputs while meeting task deadlines, thanks to opportunistic shedding of the least necessary optional parts. The work is motivated by the proliferation of increasingly ubiquitous but resource-constrained embedded devices (for applications ranging from autonomous cars to the Internet of Things) and the desire to develop services that endow them with intelligence. Experiments on recent GPU hardware and a state of the art deep neural network for machine vision illustrate that our scheme can increase the overall accuracy by 10%-20% while incurring (nearly) no deadline misses.

cs.LG↗

arXiv2020

ControlVAE: Controllable Variational Autoencoder

Variational Autoencoders (VAE) and their variants have been widely used in a variety of applications, such as dialog generation, image generation and disentangled representation learning. However, the existing VAE models have some limitations in different applications. For example, a VAE easily suffers from KL vanishing in language modeling and low reconstruction quality for disentangling. To address these issues, we propose a novel controllable variational autoencoder framework, ControlVAE, that combines a controller, inspired by automatic control theory, with the basic VAE to improve the performance of resulting generative models. Specifically, we design a new non-linear PI controller, a variant of the proportional-integral-derivative (PID) control, to automatically tune the hyperparameter (weight) added in the VAE objective using the output KL-divergence as feedback during model training. The framework is evaluated using three applications; namely, language modeling, disentangled representation learning, and image generation. The results show that ControlVAE can achieve better disentangling and reconstruction quality than the existing methods. For language modelling, it not only averts the KL-vanishing, but also improves the diversity of generated text. Finally, we also demonstrate that ControlVAE improves the reconstruction quality of generated images compared to the original VAE.

cs.LG↗

arXiv2019

STFNets: Learning Sensing Signals from the Time-Frequency Perspective with Short-Time Fourier Neural Networks

Recent advances in deep learning motivate the use of deep neural networks in Internet-of-Things (IoT) applications. These networks are modelled after signal processing in the human brain, thereby leading to significant advantages at perceptual tasks such as vision and speech recognition. IoT applications, however, often measure physical phenomena, where the underlying physics (such as inertia, wireless signal propagation, or the natural frequency of oscillation) are fundamentally a function of signal frequencies, offering better features in the frequency domain. This observation leads to a fundamental question: For IoT applications, can one develop a new brand of neural network structures that synthesize features inspired not only by the biology of human perception but also by the fundamental nature of physics? Hence, in this paper, instead of using conventional building blocks (e.g., convolutional and recurrent layers), we propose a new foundational neural network building block, the Short-Time Fourier Neural Network (STFNet). It integrates a widely-used time-frequency analysis method, the Short-Time Fourier Transform, into data processing to learn features directly in the frequency domain, where the physics of underlying phenomena leave better foot-prints. STFNets bring additional flexibility to time-frequency analysis by offering novel nonlinear learnable operations that are spectral-compatible. Moreover, STFNets show that transforming signals to a domain that is more connected to the underlying physics greatly simplifies the learning process. We demonstrate the effectiveness of STFNets with extensive experiments. STFNets significantly outperform the state-of-the-art deep learning models in all experiments. A STFNet, therefore, demonstrates superior capability as the fundamental building block of deep neural networks for IoT applications for various sensor inputs.

cs.LG↗

arXiv2018

FastDeepIoT: Towards Understanding and Optimizing Neural Network Execution Time on Mobile and Embedded Devices

Deep neural networks show great potential as solutions to many sensing application problems, but their excessive resource demand slows down execution time, pausing a serious impediment to deployment on low-end devices. To address this challenge, recent literature focused on compressing neural network size to improve performance. We show that changing neural network size does not proportionally affect performance attributes of interest, such as execution time. Rather, extreme run-time nonlinearities exist over the network configuration space. Hence, we propose a novel framework, called FastDeepIoT, that uncovers the non-linear relation between neural network structure and execution time, then exploits that understanding to find network configurations that significantly improve the trade-off between execution time and accuracy on mobile and embedded devices. FastDeepIoT makes two key contributions. First, FastDeepIoT automatically learns an accurate and highly interpretable execution time model for deep neural networks on the target device. This is done without prior knowledge of either the hardware specifications or the detailed implementation of the used deep learning library. Second, FastDeepIoT informs a compression algorithm how to minimize execution time on the profiled device without impacting accuracy. We evaluate FastDeepIoT using three different sensing-related tasks on two mobile devices: Nexus 5 and Galaxy Nexus. FastDeepIoT further reduces the neural network execution time by $48\%$ to $78\%$ and energy consumption by $37\%$ to $69\%$ compared with the state-of-the-art compression algorithms.

cs.LG↗

arXiv2014

PRINCE: Privacy-Preserving Mechanisms for Influence Diffusion in Online Social Networks

This paper has been withdrawn by the author due to a crucial sign error in equation 1. With the advance of online social networks, there has been extensive research on how to spread influence in online social networks, and many algorithms and models have been proposed. However, many fundamental problems have also been overlooked. Among those, the most important problems are the incentive aspect and the privacy aspect (eg, nodes' relationships) of the influence propagation in online social networks. Bearing these defects in mind, and incorporating the powerful tool from differential privacy, we propose PRINCE, which is a series of \underline{PR}ivacy preserving mechanisms for \underline{IN}fluen\underline{CE} diffusion in online social networks to solve the problems. We not only theoretically prove many elegant properties of PRINCE, but also implement PRINCE to evaluate its performance extensively. The evaluation results show that PRINCE achieves good performances. To the best of our knowledge, PRINCE is the first differentially private mechanism for influence diffusion in online social networks.

cs.SI↗