arXiv subjects

Guang Cheng

Publications and source records attributed to Guang Cheng.

At least 73 records · Page 4Linked to original sources

arXiv2025

PRIME: Plasticity-Robust Incremental Model for Encrypted Traffic Classification in Dynamic Network Environments

With the continuous development of network environments and technologies, ensuring cyber security and governance is increasingly challenging. Network traffic classification(ETC) can analyzes attributes such as application categories and malicious intent, supporting network management services like QoS optimization, intrusion detection, and targeted billing. As the prevalence of traffic encryption increases, deep learning models are relied upon for content-agnostic analysis of packet sequences. However, the emergence of new services and attack variants often leads to incremental tasks for ETC models. To ensure model effectiveness, incremental learning techniques are essential; however, recent studies indicate that neural networks experience declining plasticity as tasks increase. We identified plasticity issues in existing incremental learning methods across diverse traffic samples and proposed the PRIME framework. By observing the effective rank of model parameters and the proportion of inactive neurons, the PRIME architecture can appropriately increase the parameter scale when the model's plasticity deteriorates. Experiments show that in multiple encrypted traffic datasets and different category increment scenarios, the PRIME architecture performs significantly better than other incremental learning algorithms with minimal increase in parameter scale.

cs.NI↗

arXiv2025

Risk In Context: Benchmarking Privacy Leakage of Foundation Models in Synthetic Tabular Data Generation

Synthetic tabular data is essential for machine learning workflows, especially for expanding small or imbalanced datasets and enabling privacy-preserving data sharing. However, state-of-the-art generative models (GANs, VAEs, diffusion models) rely on large datasets with thousands of examples. In low-data settings, often the primary motivation for synthetic data, these models can overfit, leak sensitive records, and require frequent retraining. Recent work uses large pre-trained transformers to generate rows via in-context learning (ICL), which needs only a few seed examples and no parameter updates, avoiding retraining. But ICL repeats seed rows verbatim, introducing a new privacy risk that has only been studied in text. The severity of this risk in tabular synthesis-where a single row may identify a person-remains unclear. We address this gap with the first benchmark of three foundation models (GPT-4o-mini, LLaMA 3.3 70B, TabPFN v2) against four baselines on 35 real-world tables from health, finance, and policy. We evaluate statistical fidelity, downstream utility, and membership inference leakage. Results show foundation models consistently have the highest privacy risk. LLaMA 3.3 70B reaches up to 54 percentage points higher true-positive rate at 1% FPR than the safest baseline. GPT-4o-mini and TabPFN are also highly vulnerable. We plot the privacy-utility frontier and show that CTGAN and GPT-4o-mini offer better tradeoffs. A factorial study finds that three zero-cost prompt tweaks-small batch size, low temperature, and using summary statistics-can reduce worst-case AUC by 14 points and rare-class leakage by up to 39 points while maintaining over 90% fidelity. Our benchmark offers a practical guide for safer low-data synthesis with foundation models.

cs.LG↗

arXiv2025

Towards High Supervised Learning Utility Training Data Generation: Data Pruning and Column Reordering

Tabular data synthesis for supervised learning ('SL') model training is gaining popularity in industries such as healthcare, finance, and retail. Despite the progress made in tabular data generators, models trained with synthetic data often underperform compared to those trained with original data. This low SL utility of synthetic data stems from class imbalance exaggeration and SL data relationship overlooked by tabular generator. To address these challenges, we draw inspirations from techniques in emerging data-centric artificial intelligence and elucidate Pruning and ReOrdering ('PRRO'), a novel pipeline that integrates data-centric techniques into tabular data synthesis. PRRO incorporates data pruning to guide the table generator towards observations with high signal-to-noise ratio, ensuring that the class distribution of synthetic data closely matches that of the original data. Besides, PRRO employs a column reordering algorithm to align the data modeling structure of generators with that of SL models. These two modules enable PRRO to optimize SL utility of synthetic data. Empirical experiments on 22 public datasets show that synthetic data generated using PRRO enhances predictive performance compared to data generated without PRRO. Specifically, synthetic replacement of original data yields an average improvement of 26.74% and up to 871.46% improvement using PRRO, while synthetic appendant to original data results with PRRO-generated data results in an average improvement of 6.13% and up to 200.32%. Furthermore, experiments on six highly imbalanced datasets show that PRRO enables the generator to produce synthetic data with a class distribution that resembles the original data more closely, achieving a similarity improvement of 43%. Through PRRO, we foster a seamless integration of data synthesis to subsequent SL prediction, promoting quality and accessible data analysis.

cs.LG↗

arXiv2025

Attack Smarter: Attention-Driven Fine-Grained Webpage Fingerprinting Attacks

Website Fingerprinting (WF) attacks aim to infer which websites a user is visiting by analyzing traffic patterns, thereby compromising user anonymity. Although this technique has been demonstrated to be effective in controlled experimental environments, it remains largely limited to small-scale scenarios, typically restricted to recognizing website homepages. In practical settings, however, users frequently access multiple subpages in rapid succession, often before previous content fully loads. WebPage Fingerprinting (WPF) generalizes the WF framework to large-scale environments by modeling subpages of the same site as distinct classes. These pages often share similar page elements, resulting in lower inter-class variance in traffic features. Furthermore, we consider multi-tab browsing scenarios, in which a single trace encompasses multiple categories of webpages. This leads to overlapping traffic segments, and similar features may appear in different positions within the traffic, thereby increasing the difficulty of classification. To address these challenges, we propose an attention-driven fine-grained WPF attack, named ADWPF. Specifically, during the training phase, we apply targeted augmentation to salient regions of the traffic based on attention maps, including attention cropping and attention masking. ADWPF then extracts low-dimensional features from both the original and augmented traffic and applies self-attention modules to capture the global contextual patterns of the trace. Finally, to handle the multi-tab scenario, we employ the residual attention to generate class-specific representations of webpages occurring at different temporal positions. Extensive experiments demonstrate that the proposed method consistently surpasses state-of-the-art baselines across datasets of different scales.

cs.CR↗

arXiv2025

Breaking Distortion-free Watermarks in Large Language Models

In recent years, LLM watermarking has emerged as an attractive safeguard against AI-generated content, with promising applications in many real-world domains. However, there are growing concerns that the current LLM watermarking schemes are vulnerable to expert adversaries wishing to reverse-engineer the watermarking mechanisms. Prior work in breaking or stealing LLM watermarks mainly focuses on the distribution-modifying algorithm of Kirchenbauer et al. (2023), which perturbs the logit vector before sampling. In this work, we focus on reverse-engineering the other prominent LLM watermarking scheme, distortion-free watermarking (Kuditipudi et al. 2024), which preserves the underlying token distribution by using a hidden watermarking key sequence. We demonstrate that, even under a more sophisticated watermarking scheme, it is possible to compromise the LLM and carry out a spoofing attack, i.e. generate a large number of (potentially harmful) texts that can be attributed to the original watermarked LLM. Specifically, we propose using adaptive prompting and a sorting-based algorithm to accurately recover the underlying secret key for watermarking the LLM. Our empirical findings on LLAMA-3.1-8B-Instruct, Mistral-7B-Instruct, Gemma-7b, and OPT-125M challenge the current theoretical claims on the robustness and usability of the distortion-free watermarking techniques.

cs.CR↗

arXiv2025

M3S-UPD: Efficient Multi-Stage Self-Supervised Learning for Fine-Grained Encrypted Traffic Classification with Unknown Pattern Discovery

The growing complexity of encrypted network traffic presents dual challenges for modern network management: accurate multiclass classification of known applications and reliable detection of unknown traffic patterns. Although deep learning models show promise in controlled environments, their real-world deployment is hindered by data scarcity, concept drift, and operational constraints. This paper proposes M3S-UPD, a novel Multi-Stage Self-Supervised Unknown-aware Packet Detection framework that synergistically integrates semi-supervised learning with representation analysis. Our approach eliminates artificial segregation between classification and detection tasks through a four-phase iterative process: 1) probabilistic embedding generation, 2) clustering-based structure discovery, 3) distribution-aligned outlier identification, and 4) confidence-aware model updating. Key innovations include a self-supervised unknown detection mechanism that requires neither synthetic samples nor prior knowledge, and a continuous learning architecture that is resistant to performance degradation. Experimental results show that M3S-UPD not only outperforms existing methods on the few-shot encrypted traffic classification task, but also simultaneously achieves competitive performance on the zero-shot unknown traffic discovery task.

cs.CR↗

arXiv2025

A Probabilistic Perspective on Model Collapse

In recent years, model collapse has become a critical issue in language model training, making it essential to understand the underlying mechanisms driving this phenomenon. In this paper, we investigate recursive parametric model training from a probabilistic perspective, aiming to characterize the conditions under which model collapse occurs and, crucially, how it can be mitigated. We conceptualize the recursive training process as a random walk of the model estimate, highlighting how the sample size influences the step size and how the estimation procedure determines the direction and potential bias of the random walk. Under mild conditions, we rigorously show that progressively increasing the sample size at each training step is necessary to prevent model collapse. In particular, when the estimation is unbiased, the required growth rate follows a superlinear pattern. This rate needs to be accelerated even further in the presence of substantial estimation bias. Building on this probabilistic framework, we also investigate the probability that recursive training on synthetic data yields models that outperform those trained solely on real data. Moreover, we extend these results to general parametric model family in an asymptotic regime. Finally, we validate our theoretical results through extensive simulations and a real-world dataset.

stat.ML↗

arXiv2025

Utility Theory of Synthetic Data Generation

Synthetic data algorithms are widely employed in industries to generate artificial data for downstream learning tasks. While existing research primarily focuses on empirically evaluating utility of synthetic data, its theoretical understanding is largely lacking. This paper bridges the practice-theory gap by establishing relevant utility theory in a statistical learning framework. It considers two utility metrics: generalization and ranking of models trained on synthetic data. The former is defined as the generalization difference between models trained on synthetic and on real data. By deriving analytical bounds for this utility metric, we demonstrate that the synthetic feature distribution does not need to be similar as that of real data for ensuring comparable generalization of synthetic models, provided proper model specifications in downstream learning tasks. The latter utility metric studies the relative performance of models trained on synthetic data. In particular, we discover that the distribution of synthetic data is not necessarily similar as the real one to ensure consistent model comparison. Interestingly, consistent model comparison is still achievable even when synthetic responses are not well generated, as long as downstream models are separable by a generalization gap. Finally, extensive experiments on non-parametric models and deep neural networks have been conducted to validate these theoretical findings.

stat.ML↗

arXiv2025

Rate-Optimal Rank Aggregation with Private Pairwise Rankings

In various real-world scenarios, such as recommender systems and political surveys, pairwise rankings are commonly collected and utilized for rank aggregation to derive an overall ranking of items. However, preference rankings can reveal individuals' personal preferences, highlighting the need to protect them from exposure in downstream analysis. In this paper, we address the challenge of preserving privacy while ensuring the utility of rank aggregation based on pairwise rankings generated from a general comparison model. A common privacy protection strategy in practice is the use of the randomized response mechanism to perturb raw pairwise rankings. However, a critical challenge arises because the privatized rankings no longer adhere to the original model, resulting in significant bias in downstream rank aggregation tasks. To address this, we propose an adaptive debiasing method for rankings from the randomized response mechanism, ensuring consistent estimation of true preferences and enhancing the utility of downstream rank aggregation. Theoretically, we provide insights into the relationship between overall privacy guarantees and estimation errors in private ranking data, and establish minimax rates for estimation errors. This enables the determination of optimal privacy guarantees that balance consistency in rank aggregation with privacy protection. We also investigate convergence rates of expected ranking errors for partial and full ranking recovery, quantifying how privacy protection affects the specification of top-$K$ item sets and complete rankings. Our findings are validated through extensive simulations and a real-world application.

stat.ML↗

arXiv2025

Unleashing the Power of LLM to Infer State Machine from the Protocol Implementation

State machines are essential for enhancing protocol analysis to identify vulnerabilities. However, inferring state machines from network protocol implementations is challenging due to complex code syntax and semantics. Traditional dynamic analysis methods often miss critical state transitions due to limited coverage, while static analysis faces path explosion issues. To overcome these challenges, we introduce a novel state machine inference approach utilizing Large Language Models (LLMs), named ProtocolGPT. This method employs retrieval augmented generation technology to enhance a pre-trained model with specific knowledge from protocol implementations. Through effective prompt engineering, we accurately identify and infer state machines. To the best of our knowledge, our approach represents the first state machine inference that leverages the source code of protocol implementations. Our evaluation of six protocol implementations shows that our method achieves a precision of over 90%, outperforming the baselines by more than 30%. Furthermore, integrating our approach with protocol fuzzing improves coverage by more than 20% and uncovers two 0-day vulnerabilities compared to baseline methods.

cs.CR↗

arXiv2025

GReaTER: Generate Realistic Tabular data after data Enhancement and Reduction

Tabular data synthesis involves not only multi-table synthesis but also generating multi-modal data (e.g., strings and categories), which enables diverse knowledge synthesis. However, separating numerical and categorical data has limited the effectiveness of tabular data generation. The GReaT (Generate Realistic Tabular Data) framework uses Large Language Models (LLMs) to encode entire rows, eliminating the need to partition data types. Despite this, the framework's performance is constrained by two issues: (1) tabular data entries lack sufficient semantic meaning, limiting LLM's ability to leverage pre-trained knowledge for in-context learning, and (2) complex multi-table datasets struggle to establish effective relationships for collaboration. To address these, we propose GReaTER (Generate Realistic Tabular Data after data Enhancement and Reduction), which includes: (1) a data semantic enhancement system that improves LLM's understanding of tabular data through mapping, enabling better in-context learning, and (2) a cross-table connecting method to establish efficient relationships across complex tables. Experimental results show that GReaTER outperforms the GReaT framework.

cs.LG↗

arXiv2025

SNAKE: A Sustainable and Multi-functional Traffic Analysis System utilizing Specialized Large-Scale Models with a Mixture of Experts Architecture

The rapid advancement of internet technology has led to a surge in data transmission, making network traffic classification crucial for security and management. However, there are significant deficiencies in its efficiency for handling multiattribute analysis and its ability to expand model knowledge, making it difficult to adapt to the ever-changing network environment and complex identification requirements. To address this issue, we proposed the SNAKE (Sustainable Network Analysis with Knowledge Exploration) system, which adopts a multi-gated mixture of experts architecture to construct a multi-functional traffic classification model. The system analyzes traffic attributes at different levels through multiple expert sub-models, providing predictions for these attributes via gating and a final Tower network. Additionally, through an intelligent gating configuration, the system enables extremely fast model integration and evolution across various knowledge expansion scenarios. Its excellent compatibility allows it to continuously evolve into a multi-functional largescale model in the field of traffic analysis. Our experimental results demonstrate that the SNAKE system exhibits remarkable scalability when faced with incremental challenges in diverse traffic classification tasks. Currently, we have integrated multiple models into the system, enabling it to classify a wide range of attributes, such as encapsulation usage, application types and numerous malicious behaviors. We believe that SNAKE can pioneeringly create a sustainable and multifunctional large-scale model in the field of network traffic analysis after continuous expansion.

cs.NI↗

arXiv2025

Drift-oriented Self-evolving Encrypted Traffic Application Classification for Actual Network Environment

Encrypted traffic classification technology is a crucial decision-making information source for network management and security protection. It has the advantages of excellent response timeliness, large-scale data bearing, and cross-time-and-space analysis. The existing research on encrypted traffic classification has gradually transitioned from the closed world to the open world, and many classifier optimization and feature engineering schemes have been proposed. However, encrypted traffic classification has yet to be effectively applied to the actual network environment. The main reason is that applications on the Internet are constantly updated, including function adjustment and version change, which brings severe feature concept drift, resulting in rapid failure of the classifier. Hence, the entire model must be retrained only past very fast time, with unacceptable labeled sample constructing and model training cost. To solve this problem, we deeply study the characteristics of Internet application updates, associate them with feature concept drift, and then propose self-evolving encrypted traffic classification. We propose a feature concept drift determination method and a drift-oriented self-evolving fine-tuning method based on the Laida criterion to adapt to all applications that are likely to be updated. In the case of no exact label samples, the classifier evolves through fully fine-tuning continuously, and the time interval between two necessary retraining is greatly extended to be applied to the actual network environment. Experiments show that our approach significantly improves the classification performance of the original classifier on the following stage dataset of the following months (9\% improvement on F1-score) without any hard-to-acquire labeled sample. Under the current experimental environment, the life of the classifier is extended to more than eight months.

cs.NI↗

arXiv2024

Decentralized Sparse Linear Regression via Gradient-Tracking: Linear Convergence and Statistical Guarantees

We study sparse linear regression over a network of agents, modeled as an undirected graph and no server node. The estimation of the $s$-sparse parameter is formulated as a constrained LASSO problem wherein each agent owns a subset of the $N$ total observations. We analyze the convergence rate and statistical guarantees of a distributed projected gradient tracking-based algorithm under high-dimensional scaling, allowing the ambient dimension $d$ to grow with (and possibly exceed) the sample size $N$. Our theory shows that, under standard notions of restricted strong convexity and smoothness of the loss functions, suitable conditions on the network connectivity and algorithm tuning, the distributed algorithm converges globally at a {\it linear} rate to an estimate that is within the centralized {\it statistical precision} of the model, $O(s\log d/N)$. When $s\log d/N=o(1)$, a condition necessary for statistical consistency, an $\varepsilon$-optimal solution is attained after $\mathcal{O}(κ\log (1/\varepsilon))$ gradient computations and $O (κ/(1-ρ) \log (1/\varepsilon))$ communication rounds, where $κ$ is the restricted condition number of the loss function and $ρ$ measures the network connectivity. The computation cost matches that of the centralized projected gradient algorithm despite having data distributed; whereas the communication rounds reduce as the network connectivity improves. Overall, our study reveals interesting connections between statistical efficiency, network connectivity \& topology, and convergence rate in high dimensions.

cs.LG↗

arXiv2024

Watermarking Generative Categorical Data

In this paper, we propose a novel statistical framework for watermarking generative categorical data. Our method systematically embeds pre-agreed secret signals by splitting the data distribution into two components and modifying one distribution based on a deterministic relationship with the other, ensuring the watermark is embedded at the distribution-level. To verify the watermark, we introduce an insertion inverse algorithm and detect its presence by measuring the total variation distance between the inverse-decoded data and the original distribution. Unlike previous categorical watermarking methods, which primarily focus on embedding watermarks into a given dataset, our approach operates at the distribution-level, allowing for verification from a statistical distributional perspective. This makes it particularly well-suited for the modern paradigm of synthetic data generation, where the underlying data distribution, rather than specific data points, is of primary importance. The effectiveness of our method is demonstrated through both theoretical analysis and empirical validation.

cs.CR↗

arXiv2024

DEREC-SIMPRO: unlock Language Model benefits to advance Synthesis in Data Clean Room

Data collaboration via Data Clean Room offers value but raises privacy concerns, which can be addressed through synthetic data and multi-table synthesizers. Common multi-table synthesizers fail to perform when subjects occur repeatedly in both tables. This is an urgent yet unresolved problem, since having both tables with repeating subjects is common. To improve performance in this scenario, we present the DEREC 3-step pre-processing pipeline to generalize adaptability of multi-table synthesizers. We also introduce the SIMPRO 3-aspect evaluation metrics, which leverage conditional distribution and large-scale simultaneous hypothesis testing to provide comprehensive feedback on synthetic data fidelity at both column and table levels. Results show that using DEREC improves fidelity, and multi-table synthesizers outperform single-table counterparts in collaboration settings. Together, the DEREC-SIMPRO pipeline offers a robust solution for generalizing data collaboration, promoting a more efficient, data-driven society.

cs.DB↗

arXiv2024

Transfer Learning for Diffusion Models

Diffusion models, a specific type of generative model, have achieved unprecedented performance in recent years and consistently produce high-quality synthetic samples. A critical prerequisite for their notable success lies in the presence of a substantial number of training samples, which can be impractical in real-world applications due to high collection costs or associated risks. Consequently, various finetuning and regularization approaches have been proposed to transfer knowledge from existing pre-trained models to specific target domains with limited data. This paper introduces the Transfer Guided Diffusion Process (TGDP), a novel approach distinct from conventional finetuning and regularization methods. We prove that the optimal diffusion model for the target domain integrates pre-trained diffusion models on the source domain with additional guidance from a domain classifier. We further extend TGDP to a conditional version for modeling the joint distribution of data and its corresponding labels, together with two additional regularization terms to enhance the model performance. We validate the effectiveness of TGDP on both simulated and real-world datasets.

cs.LG↗

arXiv2024

Data Deletion for Linear Regression with Noisy SGD

In the current era of big data and machine learning, it's essential to find ways to shrink the size of training dataset while preserving the training performance to improve efficiency. However, the challenge behind it includes providing practical ways to find points that can be deleted without significantly harming the training result and suffering from problems like underfitting. We therefore present the perfect deleted point problem for 1-step noisy SGD in the classical linear regression task, which aims to find the perfect deleted point in the training dataset such that the model resulted from the deleted dataset will be identical to the one trained without deleting it. We apply the so-called signal-to-noise ratio and suggest that its value is closely related to the selection of the perfect deleted point. We also implement an algorithm based on this and empirically show the effectiveness of it in a synthetic dataset. Finally we analyze the consequences of the perfect deleted point, specifically how it affects the training performance and privacy budget, therefore highlighting its potential. This research underscores the importance of data deletion and calls for urgent need for more studies in this field.

stat.ML↗