Machine Learning Street Talk (MLST) cover image

How Do AI Models Actually Think? - Laura Ruis

Machine Learning Street Talk (MLST)

00:00

Influence of Pre-Training Data on AI Reasoning

This chapter investigates how influence functions can assess the impact of pre-training data on AI model behavior, distinguishing between factual retrieval and reasoning tasks. It highlights the complexities of reasoning in AI, exploring how models generate knowledge from various sources while addressing philosophical questions about knowledge and belief. The discussion also examines the role of coding in AI reasoning and the challenges of modeling natural language, emphasizing the transition from concrete to abstract understanding in language acquisition.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app