Model Sensitivity
Aidan shares the surprising sensitivity of models to human-collected data, highlighting the critical impact of even a few errors on model direction. Lukas expresses astonishment at the delicate nature of these models, shedding light on the challenges beyond just training them.In this clip
From this podcast

Gradient Dissent - A Machine Learning Podcast
Scaling LLMs and Accelerating Adoption: Interview with Aidan Gomez
Related Questions