Language Model Scaling
Stella and Lukas discuss the implications of training and releasing very large language models, emphasizing the importance of studying their properties and limitations. They question the necessity of continuously scaling models and highlight the focus on interpretability and alignment in current research efforts.In this clip
From this podcast

Gradient Dissent - A Machine Learning Podcast
How EleutherAI Trains and Releases LLMs: Interview with Stella Biderman
Related Questions