arXiv subjects

Yizhong Ding

Publications and source records attributed to Yizhong Ding.

6 recordsLinked to original sources

arXiv2026

Membership Inference for Contrastive Pre-training Models with Text-only PII Queries

Contrastive pretraining models such as CLIP and CLAP, serve as the ubiquitous perceptual backbones for modern multimodal large models, yet their reliance on web-scale data raises growing concerns about memorizing Personally Identifiable Information (PII). Auditing such models via membership inference is challenging in practice: shadow-model MIAs are computationally prohibitive for large multimodal backbones, and existing multimodal auditing methods typically require querying the target with paired biometric inputs, thereby directly exposing sensitive biometric information to the target model. To bypass this critical limitation, we demonstrate a highly desirable capability for privacy auditing: multimodal memorization within these foundational encoders can be accurately inferred using exclusively the text modality. We propose Unimodal Membership Inference Detector (UMID), a text-only auditing framework that performs text-guided cross-modal latent inversion and extracts two complementary signals, similarity (alignment to the queried text) and variability (consistency across randomized inversions). UMID compares these statistics to a lightweight non-member reference constructed from synthetic gibberish and makes decisions via an ensemble of unsupervised anomaly detectors. Comprehensive experiments across diverse CLIP and CLAP architectures demonstrate that UMID significantly improves the effectiveness and efficiency over prior MIAs, delivering strong detection performance with sub-second auditing cost using solely text queries, completely circumventing the need for biometric inputs and complying with strict privacy constraints.

cs.CR↗

arXiv2026

ShellForge: Adversarial Co-Evolution of Webshell Generation and Multi-View Detection for Robust Webshell Defense

Webshells remain a primary foothold for attackers to compromise servers, particularly within PHP ecosystems. However, existing detection mechanisms often struggle to keep pace with rapid variant evolution and sophisticated obfuscation techniques that camouflage malicious intent. Furthermore, many current defenses suffer from high false-alarm rates when encountering benign administrative scripts that employ heavy obfuscation for intellectual property protection. To address these challenges, we present ShellForge, an adversarial co-evolution framework that couples automated webshell generation with multi-view detection to continuously harden defensive boundaries. The framework operates through an iterative co-training loop where a generator and a detector mutually reinforce each other via the exchange of hard samples. The generator is optimized through supervised fine-tuning and preference-based reinforcement learning to synthesize functional, highly evasive variants. Simultaneously, we develop a multi-view fusion detector that integrates semantic features from long-string compression, structural features from pruned abstract syntax trees, and global statistical indicators such as Shannon entropy. To minimize false positives, ShellForge utilizes a LLM-based transformation to create de-malicious samples--scripts that retain complex obfuscation patterns but lack harmful payloads--serving as high-quality hard negatives during training. Evaluations on the public FWOID benchmark demonstrate that ShellForge significantly enhances defensive robustness. Upon convergence, the detector maintains a 0.981 F1-score while the generator achieves a 0.939 evasion rate against commercial engines on VirusTotal.

cs.CR↗

arXiv2026

AutoVulnPHP: LLM-Powered Two-Stage PHP Vulnerability Detection and Automated Localization

PHP's dominance in web development is undermined by security challenges: static analysis lacks semantic depth, causing high false positives; dynamic analysis is computationally expensive; and automated vulnerability localization suffers from coarse granularity and imprecise context. Additionally, the absence of large-scale PHP vulnerability datasets and fragmented toolchains hinder real-world deployment. We present AutoVulnPHP, an end-to-end framework coupling two-stage vulnerability detection with fine-grained automated localization. SIFT-VulMiner (Structural Inference for Flaw Triage Vulnerability Miner) generates vulnerability hypotheses using AST structures enhanced with data flow. SAFE-VulMiner (Semantic Analysis for Flaw Evaluation Vulnerability Miner) verifies candidates through pretrained code encoder embeddings, eliminating false positives. ISAL (Incremental Sequence Analysis for Localization) pinpoints root causes via syntax-guided tracing, chain-of-thought LLM inference, and causal consistency checks to ensure precision. We contribute PHPVD, the first large-scale PHP vulnerability dataset with 26,614 files (5.2M LOC) across seven vulnerability types. On public benchmarks and PHPVD, AutoVulnPHP achieves 99.7% detection accuracy, 99.5% F1 score, and 81.0% localization rate. Deployed on real-world repositories, it discovered 429 previously unknown vulnerabilities, 351 assigned CVE identifiers, validating its practical effectiveness.

cs.CR↗

arXiv2025

PBI-Attack: Prior-Guided Bimodal Interactive Black-Box Jailbreak Attack for Toxicity Maximization

Understanding the vulnerabilities of Large Vision Language Models (LVLMs) to jailbreak attacks is essential for their responsible real-world deployment. Most previous work requires access to model gradients, or is based on human knowledge (prompt engineering) to complete jailbreak, and they hardly consider the interaction of images and text, resulting in inability to jailbreak in black box scenarios or poor performance. To overcome these limitations, we propose a Prior-Guided Bimodal Interactive Black-Box Jailbreak Attack for toxicity maximization, referred to as PBI-Attack. Our method begins by extracting malicious features from a harmful corpus using an alternative LVLM and embedding these features into a benign image as prior information. Subsequently, we enhance these features through bidirectional cross-modal interaction optimization, which iteratively optimizes the bimodal perturbations in an alternating manner through greedy search, aiming to maximize the toxicity of the generated response. The toxicity level is quantified using a well-trained evaluation model. Experiments demonstrate that PBI-Attack outperforms previous state-of-the-art jailbreak methods, achieving an average attack success rate of 92.5% across three open-source LVLMs and around 67.3% on three closed-source LVLMs. Disclaimer: This paper contains potentially disturbing and offensive content.

cs.CR↗

arXiv2025

A Reward-driven Automated Webshell Malicious-code Generator for Red-teaming

Frequent cyber-attacks have elevated WebShell exploitation and defense to a critical research focus within network security. However, there remains a significant shortage of publicly available, well-categorized malicious-code datasets organized by obfuscation method. Existing malicious-code generation methods, which primarily rely on prompt engineering, often suffer from limited diversity and high redundancy in the payloads they produce. To address these limitations, we propose \textbf{RAWG}, a \textbf{R}eward-driven \textbf{A}utomated \textbf{W}ebshell Malicious-code \textbf{G}enerator designed for red-teaming applications. Our approach begins by categorizing webshell samples from common datasets into seven distinct types of obfuscation. We then employ a large language model (LLM) to extract and normalize key tokens from each sample, creating a standardized, high-quality corpus. Using this curated dataset, we perform supervised fine-tuning (SFT) on an open-source large model to enable the generation of diverse, highly obfuscated webshell malicious payloads. To further enhance generation quality, we apply Proximal Policy Optimization (PPO), treating malicious-code samples as "chosen" data and benign code as "rejected" data during reinforcement learning. Extensive experiments demonstrate that RAWG significantly outperforms current state-of-the-art methods in both payload diversity and escape effectiveness.

cs.CR↗

arXiv2024

Gibberish is All You Need for Membership Inference Detection in Contrastive Language-Audio Pretraining

Audio can disclose PII, particularly when combined with related text data. Therefore, it is essential to develop tools to detect privacy leakage in Contrastive Language-Audio Pretraining(CLAP). Existing MIAs need audio as input, risking exposure of voiceprint and requiring costly shadow models. We first propose PRMID, a membership inference detector based probability ranking given by CLAP, which does not require training shadow models but still requires both audio and text of the individual as input. To address these limitations, we then propose USMID, a textual unimodal speaker-level membership inference detector, querying the target model using only text data. We randomly generate textual gibberish that are clearly not in training dataset. Then we extract feature vectors from these texts using the CLAP model and train a set of anomaly detectors on them. During inference, the feature vector of each test text is input into the anomaly detector to determine if the speaker is in the training set (anomalous) or not (normal). If available, USMID can further enhance detection by integrating real audio of the tested speaker. Extensive experiments on various CLAP model architectures and datasets demonstrate that USMID outperforms baseline methods using only text data.

cs.SD↗