2019
DOI: 10.3233/sat190123
|View full text |Cite
|
Sign up to set email alerts
|

The SMT Competition 2015–2018

Abstract: The International Satisfiability Modulo Theories Competition is an annual competition between Satisfiability Modulo Theories (SMT) solvers. The 2018 edition of the competition was part of the FLoC Olympic Games, which comprised 14 competitions in various areas of computational logic. We report on the design and selected results of the SMT Competition during the last FLoC Olympiad, from 2015 to 2018. These competitions set several new records regarding the number of participants, number of benchmarks used, and … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
12
0

Year Published

2020
2020
2022
2022

Publication Types

Select...
5
2
1

Relationship

1
7

Authors

Journals

citations
Cited by 28 publications
(12 citation statements)
references
References 12 publications
0
12
0
Order By: Relevance
“…Unless otherwise specified, all tools are run without any previous training. For consistency, all graphs were generated by evaluating the 2018-Goel-hwbench dataset [11] in the QF-BV (quantifier free bit-vector) theory.…”
Section: Models and Methodsmentioning
confidence: 99%
“…Unless otherwise specified, all tools are run without any previous training. For consistency, all graphs were generated by evaluating the 2018-Goel-hwbench dataset [11] in the QF-BV (quantifier free bit-vector) theory.…”
Section: Models and Methodsmentioning
confidence: 99%
“…Based on the list of supported theories, AUFLIRA satisfies G1. It also satisfies G2, since it is included in the SMT-LIB logic, whose satisfiability can be verified using highly efficient and optimized solvers, as shown in the annual SMT competition [23].…”
Section: Value Term Interpretationmentioning
confidence: 96%
“…In a first step we filtered out benchmarks that Vampire could solve within 1 second in both default mode (which involves a simpler version of the rule eval), and in default mode with eval enabled. Our main experiments were carried out on the remaining set of 21,512 benchmarks, we which will refer to as B. Filtering out trivial benchmarks avoids the results containing noise from benchmarks that can easily be solved and is an approach recently adopted by SMT-COMP [22]. Experiments are run on a Linux cluster where each node contains two octacore 2.1 GHz Intel Xeon processors and 160GB of RAM.…”
Section: Experimental Evaluationmentioning
confidence: 99%