The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) cover image

Scaling Up Test-Time Compute with Latent Reasoning with Jonas Geiping - #723

The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)

CHAPTER

Visualizing Recurrent States in Machine Learning

This chapter explores the evolution of random states in machine learning models as they predict tokens, illustrating the unexpected behaviors that emerge during this process. It also examines the relationship between recurrent and diffusion models, highlighting both their conceptual similarities and distinct training challenges in language processing.

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner