7 papers
Reveree: Diagnosing LLM Reverse-Engineering Agents
Hadjer Benkraouda, Hongyu Cai, Berkay Celik +1
Reverse engineering (RE) is critical to security tasks such as malware analysis and vulnerability discovery, and large language model (LLM) agents are increasingly able to perform…
Words Speak Louder Than Code: Investigating Cognitive Heuristics in LLM-Based Code Vulnerability Detection
Asif Shahriar, Hongyu Cai, Hadjer Benkraouda +2
Researchers and practitioners increasingly apply Large Language Models (LLMs) for automated vulnerability detection. Recent work has shown that LLMs are susceptible to the same cog…
RoboJailBench: Benchmarking Adversarial Attacks and Defenses in Embodied Robotic Agents
Doguhuan Yeke, Yanming Zhou, Leo Y. Lin +3
Recent advances in Vision-Language Models (VLMs) facilitate a new class of embodied AI systems, where these models are integrated into physical platforms, e.g. robots and autonomou…
Exploring and Developing a Pre-Model Safeguard with Draft Models
Hongyu Cai, Arjun Arunasalam, Yiming Liang +2
Large Language Model (LLM) alignment remains vulnerable to jailbreak attacks that elicit unsafe responses, motivating pre-model and post-model guards. Pre-model guards audit the sa…
D-LiFT: Improving LLM-based Decompiler Backend via Code Quality-driven Fine-tuning
Muqi Zou, Hongyu Cai, Hongwei Wu +9
As one of the key tools in many security tasks, decompilers reconstruct human-readable source code from binaries. Yet, despite recent advances, their outputs often suffer from synt…
A Progressive Transformer for Unifying Binary Code Embedding and Knowledge Transfer
Hanxiao Lu, Hongyu Cai, Yiming Liang +2
Language model approaches have recently been integrated into binary analysis tasks, such as function similarity detection and function signature recovery. These models typically em…
Rethinking How to Evaluate Language Model Jailbreak
Hongyu Cai, Arjun Arunasalam, Leo Y. Lin +2
Large language models (LLMs) have become increasingly integrated with various applications. To ensure that LLMs do not generate unsafe responses, they are aligned with safeguards t…