Story
“Next-token predictor” is the wrong mental model for LLMs
1 report
1 website
1 discussion
Latest report
Overview
Strictly speaking, the statement “LLMs are next-token predictors” isn’t wrong, but it’s incomplete. It’s a fine zeroth-order approximation, and it is grounded in something real: transformer-based language models emit tokens autoregressively: while not done: tokens.append(model.sample_next_token(tokens)) This certainly has the shape of something you might call a next-token predictor. During pre-tra...
Summary from gmcgoldr.github.io.
Reports
One report so far. Additional coverage will appear here when it is linked to this story.
gmcgoldr.github.io
“Next-token predictor” is the wrong mental model for LLMs
Read report →
Community reactions
Each reaction summarizes sampled comments from the linked discussion.
Hacker News · gmcgoldr.github.io
Historical count: 15 · 15 saved in sample · Historical sample
Awaiting analysis