Adversarial Code Obfuscation for Defending Against LLM-Based Analysis10 hours agohttps://arxiv.org/abs/2606.11755