English

Improved Bounds on Access-Redundancy Tradeoffs in Quantized Linear Computations

Information Theory 2025-08-04 v1 math.IT

Abstract

Consider the problem of computing quantized linear functions with only a few queries. Formally, given xRk\mathbf{x}\in \mathbb{R}^k, our goal is to encode x\mathbf{x} as cRn\mathbf{c} \in \mathbb{R}^n, for n>kn > k, so that for any wAk\mathbf{w} \in A^k, wTx\mathbf{w}^T \mathbf{x} can be computed using at most \ell queries to c\mathbf{c}. Here, AA is some finite set; in this paper we focus on the case where A=2|A| = 2. Prior work \emph{(Ramkumar, Raviv, and Tamo, Trans. IT, 2024)} has given constructions and established impossibility results for this problem. We give improved impossibility results, both for the general problem, and for the specific class of construction (block construction) presented in that work. The latter establishes that the block constructions of prior work are optimal within that class. We also initiate the study of \emph{approximate} recovery for this problem, where the goal is not to recover wTx\mathbf{w}^T \mathbf{x} exactly but rather to approximate it up to a parameter ε>0\varepsilon > 0. We give several constructions, and give constructions for ε=0.1\varepsilon = 0.1 that outperform our impossibility result for exact schemes.

Keywords

Cite

@article{arxiv.2508.00183,
  title  = {Improved Bounds on Access-Redundancy Tradeoffs in Quantized Linear Computations},
  author = {Ching-Fang Li and Mary Wootters},
  journal= {arXiv preprint arXiv:2508.00183},
  year   = {2025}
}

Comments

ISIT 2025

R2 v1 2026-07-01T04:28:37.553Z