Anthropic’s Claude Fable 5.2 Beats GPT-6 Cyber in Multi-Stage Cyberattack Benchmark

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

Leave a Reply

Your email address will not be published. Required fields are marked *