Know What's Happening in Technology

Small Models Also Found the Vulnerabilities that Mythos Found

Researchers tested a suite of inexpensive, open‑weight language models on the same code snippets Anthropic highlighted for its Mythos system. All eight small models flagged Mythos's flagship FreeBSD exploit, including a 3.6 billion‑parameter model that costs roughly $0.11 per million tokens. A 5.1 billion‑parameter model also reconstructed the core chain of a 27‑year‑old OpenBSD vulnerability. While the findings prove that modest models can spot the same bugs, commentators note they generate more false positives and lack Mythos’s ability to chain exploits into functional attacks.