Thread by Cameron R. Wolfe on Twitter about why GPT-3 is better than larger language models
Posted by jpluimers on 2025/09/04
For my link archive the [Wayback/Archive] Thread by @cwolferesearch on Thread Reader App starting with [Wayback/Archive] Cameron R. Wolfe on Twitter: “After GPT-3 was proposed, a lot of research was done to find an even better language model. Initial attempts focused on just training larger models. Contrary to popular belief, however, there is more to creating a good language model than size… 🧵[1/8]” / Twitter
3 years later, I’m anxious to know what the current state of the art on GPT is, as between GPT-2 and GPT-3 there was about a 3 year period.
–jeroen
The Wiert Corner - irregular stream of stuff 
Leave a comment