Controllability Insights
Keith shares groundbreaking findings on the relationship between control input tokens and output reachability in machine learning systems. Through empirical experiments, he demonstrates the ability to steer models towards correct outputs with minimal control input, shedding light on the controllability of complex systems.In this clip
From this podcast
Machine Learning Street Talk (MLST)
What’s the Magic Word? A Control Theory of LLM Prompting.
Related Questions