Fable/Mythos
More details on Fable 5’s cyber safeguards and our jailbreak framework \ Anthropic
https://www.anthropic.com/news/fable-safeguards-jailbreak-framework
What is and isn't blocked by our cyber classifiers, and a first draft of our jailbreak severity framework
Added 1 month ago
Shepherd's Dog: A Game by the World's Most Dangerous AI | Koen van Gilst
https://koenvangilst.nl/lab/claude-fable-shepherds-dog
A few days ago Anthropic released a model that was initially too dangerous for the world. I tested it with my personal benchmark - can it create a game idea I've had for years in one shot?