Defend LLMs Through Self-Consciousness

This research introduces a novel self-consciousness framework utilizing Meta-Cognitive and Arbitration modules to autonomously defend Large Language Models a...

Level: advanced

By Unknown

Category: discussion