Modifying LLM Beliefs with Synthetic Document Finetuning
Shows synthetic document finetuning can insert beliefs into LLMs, and demonstrates applications to honeypotting for misalignment detection and unlearning of hazardous knowledge.
Research
Anthropic
Research on reliable, interpretable and steerable AI. anthropic.com
Focus All
Topic Deception + Unlearning
Source Anthropic · 1
Date Any time
1 item