Claude Mythos: Anthropic Bans AI That Hacked OpenBSD and FFmpeg
Anthropic blocked public access to Claude Mythos, a model that found bugs hidden for 27 years. The U.S. Treasury convened an emergency meeting.
3 articles found
Anthropic blocked public access to Claude Mythos, a model that found bugs hidden for 27 years. The U.S. Treasury convened an emergency meeting.
OpenAI announces a new Instant Mini model and a $100/month Pro plan with expanded Codex access. We compare the 4 current ChatGPT tiers.
Google released TurboQuant, a technique that shrinks the KV cache via PolarQuant rotation and makes large models run on mobile devices.