Sol Searching | Can Frontier Models Tackle Autonomous Long-Horizon Malware Analysis?

A real-world benchmark tests whether powerful AI models can keep an investigation trustworthy when new evidence invalidates their conclusions.

≈