#LLM

CyberSOCEval: A Surprising Lesson from the Malware Reasoning Benchmark