RC RANDOM CHAOS

LLM deployment

3 posts

Parameter count is the wrong number to watch
Article

Parameter count is the wrong number to watch

Bonsai 2 27B's 9x smaller footprint is real, but near-lossless is an average - how to gate compressed LLMs on your own eval set before production.

The same AI you're shipping wrote the malware
Article

The same AI you're shipping wrote the malware

10,000 trojan GitHub repos weren't a malware breakthrough - they prove LLM safety lives in the model while abuse happens in the unguarded pipeline.

Forge guardrails took an 8B model from 53% to 99%
Article

Forge guardrails took an 8B model from 53% to 99%

A Show HN post says Forge guardrails took an 8B model from 53% to 99% on agentic tasks. Here's what that means for security and reliability.