In their seminal work, Atserias et al. and independently Pipatsrisawat and Darwiche in 2009 showed that CDCL solvers can simulate resolution proofs with polynomial overhead. However, previous work does not address the tightness of the simulation, i.e., the question of how large this overhead needs to be. In this paper, we address this question by focusing on an important property of proofs generated by CDCL solvers that employ standard learning schemes, namely that the derivation of a learned clause has at least one inference where a literal appears in both premises (aka, a merge literal). Specifically, we show that proofs of this kind can simulate resolution proofs with at most a linear overhead, but there also exist formulas where such overhead is necessary or, more precisely, that there exist formulas with resolution proofs of linear length that require quadratic CDCL proofs.
翻译:CDCL 的学习能力限制:合并消解
在经典学术文章中Atserias 等人和 Pipatsrisawat 与 Darwiche 2009 年分别独立地表明,CDCL 求解器可以带有多项式复杂度的开销模拟 原始合取式。 但是,过去的研究并未涉及模拟的紧密程度,即其开销需要多大。在本文中,我们将关注CDCL 求解器生成证明的重要性质,即通过标准学习方案派生学习公式的证明至少具有一个推断,其中文字出现在两个前提中。为此,我们证明了此类证明最多可以模拟具有线性开销的原始合取式,但同时也存在公式,其合并证明长度是线性的,但需要用二次 CDCL 证明才能得到正确解。