
Explore how PentestGPT leverages LLMs to automate penetration testing, addressing challenges in context maintenance and demonstrating significant improvements in task completion for security testing.

Explore a systematic framework for understanding and evaluating prompt injection attacks on LLM applications, including analysis of 5 attack types and 10 defense mechanisms across multiple models.

Discover a novel framework for detecting security vulnerabilities in RESTful APIs through functionality-based testing, featuring practical demonstrations and real-world applications in API security testing.

Explore advanced vulnerability detection techniques using multi-dimensional neighbor embeddings and deep learning models to enhance software security through improved pattern recognition and data analysis.

Discover how FIRE combines multi-stage filtering and taint analysis to efficiently detect recurring vulnerabilities in open-source software, achieving 90% precision and superior performance over existing tools.

Explore the capabilities and limitations of Large Language Models in code analysis, focusing on their effectiveness with obfuscated code and real-world applications through systematic evaluation.

Discover how automated detection frameworks identify security flaws in web browser mechanisms, exploring practical approaches to validate web invariants and detect vulnerabilities across major browsers.

Explore the challenges and solutions of binary-level thread sanitization through BINTSAN, a novel approach for detecting data races in Linux x86-64 targets without source code access.

Explore FrameFlip, a groundbreaking attack method that depletes DNN model inference through single-bit fault injections, demonstrating universal effectiveness across major machine learning frameworks.

Explore a comprehensive analysis of C decompiler output fidelity issues, examining defects beyond missing names and types to improve reverse engineering effectiveness in security analysis.

Discover a groundbreaking approach to hypervisor security testing through HYPERPILL - a universal fuzzing framework that identifies vulnerabilities across major hypervisors using hardware virtualization interfaces.

Discover how CONCH improves null pointer dereference fixes by incorporating contextual checks, achieving 85%+ patch accuracy through innovative intraprocedural and interprocedural analysis techniques.