We're publishing our most detailed threat intelligence report to date.
It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them.
We disrupted every operation in the report,
Looking a bit into local AI these days and ended up with an usual setup combining @AMD Ryzen 395+ (Geekom A9 Mega 128GB) with a @RadeonPRO R9700 32GB. Still figuring things out but here are some details:
Today we're releasing Laguna S 2.1, our most capable model to date.
It's a 118B total parameter Mixture-of-Experts model with 8B activated per token, a context window of up to 1M tokens, and thinking and no-thinking modes.
Capable enough to hold its own against models many
This helped a bit to get fable going (.claude/settings.local.json):
{
"env": {
"CLAUDE_CODE_DISABLE_AUTO_MEMORY": "1",
"CLAUDE_CODE_DISABLE_GIT_INSTRUCTIONS": "1"
}
}
