Prompting for Causality
Robert discusses the challenges faced by AI models in causal discovery tasks, highlighting how subtle prompt adjustments can lead to improved performance. While the models demonstrate coherent reasoning, they often struggle with logical errors and lack the ability to ask clarifying questions. There's potential for better outcomes through enhanced prompting strategies, though this area remains underexplored.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Are LLMs Good at Causal Reasoning? with Robert Osazuwa Ness - 638
Related Questions