
RAG Risks: Why Retrieval-Augmented LLMs are Not Safer with Sebastian Gehrmann - #732
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
00:00
Exploring Prompt Modifications for Enhanced Safety in RAG Systems
This chapter examines how altering prompts can mitigate unsafe outputs in Retrieval-Augmented Generation systems. It highlights the potential of tailored prompting strategies to reduce harmful responses, despite the lack of direct testing on this concept.
Transcript
Play full episode