:
[AI]■ STORY TIMELINE

MIT REVEALS WHY SCALING LANGUAGE MODELS WORKS

MIT researchers have identified the mechanistic reason behind language models' reliable performance improvements as they grow larger: a phenomenon called superposition.

1 SOURCEFIRST SEEN MAY 3, 08:42 AM► READ THE ARTICLE
The Decoder+0m

MIT researchers have a mechanistic explanation for why large language model performance scales so reliably with size. Th…