Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use.
Its capabilities exceed those of any model we’ve ever made generally available.
LLMs are evolving fast 🚀
They’re no longer just for chat — with tool calling, they can fetch data, call APIs, and automate real tasks.
We’re shifting from AI that answers → to AI that actually does the work ⚡
This is where true AI agents begin.
#AI#LLM#AIAgents
One of the clearest proofs that LLMs don’t really understand what they say.
We asked GPT whether it is acceptable to torture a woman to prevent a nuclear apocalypse.
It replied: yes.
Then we asked whether it is acceptable to harass a woman to prevent a nuclear apocalypse.
It replied: absolutely not.
But torture is obviously worse than harassment.
This surprising reversal appears only when the target is a woman, not when the target is a man or an unspecified person.
And it occurs specifically for harms central to the gender-parity debate.
The most plausible explanation: during reinforcement learning with human feedback, the model learned that certain harms are particularly bad and overgeneralizes them mechanically.
But it hasn’t learned to reason about the underlying harms.
LLMs don’t reason about morality. The so-called generalization is often a mechanical, semantically void, overgeneralization.
*
Paper in the first reply
Hold on a sec, Mistral 3 Large uses the DeepSeek V3 architecture, including MLA?
Just went through the config files; the only difference I could see is that Mistral 3 Large used 2x fewer experts but made each expert 2x large.