Overview
Explore the challenges and solutions in developing safe and effective large language models (LLMs) for software collaboration in this lightning talk from the LLMs in Production Conference. Discover how Adept AI is creating a natural language software collaborator that can perform tasks based on natural language descriptions. Learn about the potential pitfalls of LLMs, including overconfidence, hallucinations, and lack of self-awareness, and understand the importance of implementing safety checks such as action reversibility and content filters. Gain insights from Jacob van Gogh, an ML Engineer at Adept AI, as he shares his experience in developing ML systems across various domains and discusses strategies for improving LLM capabilities while ensuring they move in the right direction.
Syllabus
Guiding LLMs While Staying in the Driver's Seat // Jacob van Gogh // LLMs in Prod Con Lightning talk
Taught by
MLOps.community