importantSYS.SOURCE: The Hacker News• 2026-11-10T13:46:14+05:30
Anthropic's Claude Fable 5.2 and Claude Mythos 5.2 Outperform GPT-6 Cyber in Cyberattack Benchmarks
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, enhancing coding, reasoning, and vision capabilities. The models assisted in software migrations, designed electronic devices, and solved physics problems during pre-release training.
Anthropic on Monday released Claude Fable 5.2 and Claude Mythos 5.2 models, which it said exhibit improvements in coding, reasoning, and vision.
During pre-release training, the large language models (LLMs) helped its engineers "complete otherwise-infeasible software migrations, designed three working electronic devices from scratch, solved two open problems in the physics of black holes, made
*** END OF TRANSMISSION ***