Expert Prompting Insights
Riley discusses the evolving effectiveness of instructing models as experts, noting that its impact has diminished with advancements in instruction tuning. He highlights the complexities of K-shot prompting, which can lead to confusion and information bleed between examples, particularly in general question answering scenarios. The conversation delves into the nuances of model understanding and the importance of context in translations.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Mental Models for Advanced ChatGPT Prompting with Riley Goodside - 652
Related Questions