Differentially private language models obey scaling laws in which compute-optimal models are roughly 10-50x smaller than non-private Chinchilla-optimal models, with large batch sizes and rapid saturation of compute.
The crossroads of innovation and privacy: Private synthetic data for generative AI
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Scaling Laws for Differentially Private Language Models
Differentially private language models obey scaling laws in which compute-optimal models are roughly 10-50x smaller than non-private Chinchilla-optimal models, with large batch sizes and rapid saturation of compute.