Learning from Few Samples: A Novel Approach for High-Quality Malcode Generation

2025/08/25 by Ma, Haijian, Liu, Daizong, Cai, Xiaowen +2
#Artificial Intelligence (cs.AI) #Cryptography and Security (cs.CR) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2508.18148

Abstract

Intrusion Detection Systems (IDS) play a crucial role in network security defense. However, a significant challenge for IDS in training detection models is the shortage of adequately labeled malicious samples. To address these issues, this paper introduces a novel semi-supervised framework GANGRL-LLM, which integrates Generative Adversarial Networks (GANs) with Large Language Models (LLMs) to enhance malicious code generation and SQL Injection (SQLi) detection capabilities in few-sample learning scenarios. Specifically, our framework adopts a collaborative training paradigm where: (1) the GAN-based discriminator improves malicious pattern recognition through adversarial learning with generated samples and limited real samples; and (2) the LLM-based generator refines the quality of malicious code synthesis using reward signals from the discriminator. The experimental results demonstrate that even with a limited number of labeled samples, our training framework is highly effective in enhancing both malicious code generation and detection capabilities. This dual enhancement capability offers a promising solution for developing adaptive defense systems capable of countering evolving cyber threats.

Citations

• Multi-Pair Temporal Sentence Grounding via Multi-Thread Knowledge Transfer Network
• Your Data Is Not Perfect: Towards Cross-Domain Out-of-Distribution Detection in Class-Imbalanced Data
• Uncertainty-Guided Appearance-Motion Association Network for Out-of-Distribution Action Detection
• A Survey on Large Language Models for Code Generation
• On the Effectiveness of Large Language Models in Domain-Specific Code Generation
• 3DHacker: Spectrum-based Decision Boundary Generation for Hard-label 3D Point Cloud Attack
• You Can Ground Earlier than See: An Effective and Efficient Pipeline for Temporal Sentence Grounding in Compressed Videos
• The Enemy of My Enemy is My Friend: Exploring Inverse Adversaries for Improving Adversarial Training
• Multi-Modal Cross-Domain Alignment Network for Video Moment Retrieval
• Restricted Black-box Adversarial Attack Against DeepFake Face Swapping
• Exploring the Devil in Graph Spectral Domain for 3D Point Cloud Attacks
• Generative Adversarial Networks
• Context-aware Biaffine Localizing Network for Temporal Sentence Grounding
• Double Self-weighted Multi-view Clustering via Adaptive View Fusion
• Learning to summarize from human feedback
• Jointly Cross- and Self-Modal Graph Attention Network for Query-Based Moment Localization
• MixMatch: A Holistic Approach to Semi-Supervised Learning
• Long Text Generation via Adversarial Training with Leaked Information
• Maximum-Likelihood Augmented Discrete Generative Adversarial Networks
• SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient
• Improved Techniques for Training GANs
• BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Related

• Single-Shot Black-Box Adversarial Attacks Against Malware Detectors: A Causal Language Model Approach
• XGen-Q: An Explainable Domain-Adaptive LLM Framework with Retrieval-Augmented Generation for Software Security
• Generative AI-Based Effective Malware Detection for Embedded Computing Systems
• Generative Adversarial Networks for Malware Detection: a Survey
• Generating Adversarial Malware Examples for Black-Box Attacks Based on GAN