Research on how knowledge is stored in large language model parameters is beginning to understand why models must be so large; Stable Diffusion 1 has only 890 million parameters (200x smaller than GPT-3's 175 billion) because it doesn't need to internalize knowledge of individual entities.
factualpending
Speaker
Knowledge graph expertEvidence Quote
“Stable Diffusion 1... How many parameters do you think are in Stable Diffusion 1? 890 million parameters... GPT-3 is about 200 times larger than Stable Diffusion 1.”
Created: 8/12/2026, 6:34:12 PM
Your notes
Loading notes...