PaperKaplan Scaling Law论文
Scaling Laws for Neural Language Models
A seminal paper published by OpenAI in 2020 that systematically describes, for the first time, the scaling relationships between neural language model performance and model parameter count, dataset size, and compute, using power-law formulations.
Core Facts
Timeline (last 90 days)
Aug 24
The Scaling Law fitted by Kaplan et al. in 2020 concluded that parameter growth should be faster than data growth, at a ratio of approximately 2.7:1.
Unverified90%
All Facts (2)
Verified
Scaling Laws were first systematically proposed and quantified by OpenAI researchers in the 2020 paper "Scaling Laws for Neural Language Models."
90%UnverifiedThe Scaling Law fitted by Kaplan et al. in 2020 concluded that parameter growth should be faster than data growth, at a ratio of approximately 2.7:1.
90%