English

Catalytic Computing and Register Programs Beyond Log-Depth

Computational Complexity 2025-04-25 v1

Abstract

In a seminal work, Buhrman et al. (STOC 2014) defined the class CSPACE(s,c)CSPACE(s,c) of problems solvable in space ss with an additional catalytic tape of size cc, which is a tape whose initial content must be restored at the end of the computation. They showed that uniform TC1TC^1 circuits are computable in catalytic logspace, i.e., CL=CSPACE(O(logn),2O(logn))CL=CSPACE(O(\log{n}), 2^{O(\log{n})}), thus giving strong evidence that catalytic space gives LL strict additional power. Their study focuses on an arithmetic model called register programs, which has been a focal point in development since then. Understanding CLCL remains a major open problem, as TC1TC^1 remains the most powerful containment to date. In this work, we study the power of catalytic space and register programs to compute circuits of larger depth. Using register programs, we show that for every ϵ>0\epsilon > 0, SAC2CSPACE(O(log2nloglogn),2O(log1+ϵn))SAC^2 \subseteq CSPACE\left(O\left(\frac{\log^2{n}}{\log\log{n}}\right), 2^{O(\log^{1+\epsilon} n)}\right) This is an O(loglogn)O(\log \log n) factor improvement on the free space needed to compute SAC2SAC^2, which can be accomplished with near-polynomial catalytic space. We also exhibit non-trivial register programs for matrix powering, which is a further step towards showing NC2CLNC^2 \subseteq CL.

Cite

@article{arxiv.2504.17412,
  title  = {Catalytic Computing and Register Programs Beyond Log-Depth},
  author = {Yaroslav Alekseev and Yuval Filmus and Ian Mertz and Alexander Smal and Antoine Vinciguerra},
  journal= {arXiv preprint arXiv:2504.17412},
  year   = {2025}
}
R2 v1 2026-06-28T23:09:40.744Z