5min chapter

Machine Learning Street Talk (MLST) cover image

#107 - Dr. RAPHAËL MILLIÈRE - Linguistics, Theory of Mind, Grounding

Machine Learning Street Talk (MLST)

CHAPTER

The Inferential Concept of Semantic Competence in AI Models

In classical symbolic models, think of SHRDOU from Terri Winograd. It's a system that could receive natural language instructions and perform operations in a virtual world called the blocks world. CHRDOU has this kind of referential semantic competence but it can't do things like reasoning about analogies. Bicom trust language models, because they're trained on this testicle learning objective. They learn complex patterns of co-occurrence statistics in language. This enables them to induce a lot of information about how words relate to each other within language based on their use in various ways.

00:00

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode