Aligning LLMs at inference time by suppressing internal concepts

(guidelabs.ai)

2 points | by adebayoj 8 hours ago ago