cs.LGAug 4, 2026

LaPrune: Controllable Differentiable Sparsity at Million Scale

Authors: Jakub Antczak, Joanna Wojciechowicz, Łukasz Struski, Jacek Tabor

Abstract

Top-kk selection determines which components of a sparse model remain active. Hard selection blocks gradients, while continuous relaxations often couple mask hardness to the selected mass. We introduce LaPrune, a mathematically exact-budget differentiable layer that controls the normalized second moment while preserving the selected mass. A LapSum barrier preserves the selection mass, and a normalized second-moment constraint moves the mask from a dense equal-mass allocation toward hard top-kk at each budget. We derive a population prediction of the saturated fraction, a near-binary limiting law, and a tight worst-case guarantee on the near-zero fraction. The normalized hardness parameter is invariant to score scale, while a fixed LapSum temperature is not.

Explore similar work

CardsList
  1. Fast LapSum: Exact Differentiable Top-k at Million Scale

    Aug 7, 2026Łukasz Struski, Joanna Wojciechowicz, Jakub Antczak +3Top-KSparsity

  2. Hyperflux: Pruning Reveals Importance

    Apr 6, 2025Eugen Barbulescu, Antonio Alexoaie, Lucian BusoniuHeterogeneous Pruning ScoresPruning