On the Dichotomy Between Privacy and Traceability in $\ell_p$ Stochastic Convex Optimization

In this paper, we investigate the necessity of memorization in stochastic convex optimization (SCO) under $\ell_p$ geometries. Informally, we say a learning algorithm memorizes $m$ samples (or is $m$-traceable) if, by analyzing its output, it is possible to identify at least $m$ of its training samples. Our main results uncover a fundamental tradeoff between traceability and excess risk in SCO. For every $p\in [1,\infty)$, we establish the existence of a risk threshold below which any sample-efficient learner must memorize a \em{constant fraction} of its sample. For $p\in [1,2]$, this threshold coincides with best risk of differentially private (DP) algorithms, i.e., above this threshold, there are algorithms that do not memorize even a single sample. This establishes a sharp dichotomy between privacy and traceability for $p \in [1,2]$. For $p \in (2,\infty)$, this threshold instead gives novel lower bounds for DP learning, partially closing an open problem in this setup. En route of proving these results, we introduce a complexity notion we term \em{trace value} of a problem, which unifies privacy lower bounds and traceability results, and prove a sparse variant of the fingerprinting lemma.

@article{voitovych2025_2502.17384,
  title={ On Traceability in $\ell_p$ Stochastic Convex Optimization },
  author={ Sasha Voitovych and Mahdi Haghifam and Idan Attias and Gintare Karolina Dziugaite and Roi Livni and Daniel M. Roy },
  journal={arXiv preprint arXiv:2502.17384},
  year={ 2025 }
}