SAT-to-SAT: Declarative Extension of SAT Solvers with New Propagators

Authors: Tomi Janhunen, Shahab Tasharrofi, Eugenia Ternovska

AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental Results: We implement our methods on top of Glucose (Audemard and Simon 2009) and show that introducing new reasoning mechanisms hugely boosts the solving time of a problem. Thus, we show that our new paradigm to solve problems by also describing their reasoning methods outperforms the existing ground and solve paradigm.
Researcher Affiliation Academia Tomi Janhunen and Shahab Tasharrofi {tomi.janhunen,shahab.tasharrofi}@aalto.fi Helsinki Institute for Information Technology HIIT Department of Computer Science Aalto University, FI-00076 AALTO, Finland Eugenia Ternovska ter@cs.sfu.ca Computing Science Department Simon Fraser University, Burnaby, BC, Canada
Pseudocode Yes Algorithm 1 shows that SAT-to-SAT extends CDCL by introducing a new propagation handling procedure S2Sprop in the common CDCL search method.
Open Source Code No The paper states 'We have implemented SAT-to-SAT by extending GLUCOSE 3.0' and 'We also use SATGRND (Gebser et al. 2015) as our grounder,' but it does not provide a link or explicit statement for the open-source release of the SAT-to-SAT solver or the P[R] language implementation.
Open Datasets No The instances of Hamiltonian path are randomly generated using a random planar graph generator that is developed as part of LEDA (Mehlhorn and N aher 1992) library.
Dataset Splits No The paper presents 'Hamiltonian Path Instances' and their solving times in Table 1 but does not specify any training, validation, or test dataset splits.
Hardware Specification Yes Table 1 summarizes the results of running our experiments on an Ubuntu 14.04 Linux desktop with kernel version 3.13.0-57, an Intel(R) Core(TM) i5-4590 CPU running at 3.30GHz, plus 16GB of memory.
Software Dependencies Yes We have implemented SAT-to-SAT by extending GLUCOSE 3.0 (Audemard and Simon 2009). Our Glucose extension closely matches Algorithm 1 except for using fast data structures such as Bit Sets to expedite problem solving. We also use SATGRND (Gebser et al. 2015) as our grounder.
Experiment Setup Yes The time limit for each instance is set at 15 minutes and the # columns in Table 1 show the number of instances that were solved using a particular encoding in 15 minutes.