Dataset Condensation Atlas

Analysis & theory

A Theoretical Study of Dataset Distillation

Zachary Izzo, James Zou

NeurIPS 2023 Workshop · first public 2023-12

paper ↗catalogued✓ abstract read

In one paragraph

Proves existence and impossibility results for exact dataset distillation of generalized linear models: a single synthetic point can exactly reproduce a model trained on the full data for GLMs, and a size-independent construction exists for linear regression with any data-independent regularizer, but no analogous construction exists for logistic regression, and kernel regression cannot in general be distilled to recover even one model exactly.

Where it sits

BibTeX (generated; prefer the venue's official entry)
@article{izzo2023theoretical,
  title   = {A Theoretical Study of Dataset Distillation},
  author  = {Zachary Izzo and James Zou},
  journal = {NeurIPS 2023 Workshop},
  year    = {2023}
}