Open fhoushmand opened 3 years ago
Thanks for the benchmark. Set comprehension should generally be thought of as a quantified formula, so cvc5 is expected to be incomplete in general when comprehension is present. I'm marking this "performance" and will look into this in more detail.
Thanks!
Hi, the input.zip consists of three simple assertions with no quantifiers (theory of sets and uf). CVC4 produces
unknown
even with the following options. I have checked cvc5/cvc5#6274 and tried bothfinite-model-find
andfull-saturate-quant
options with no luck. Are there other options to try? Because it seems to me that this example should not hit the limit of CVC4.Command line arguments: cvc4 --sets-ext --finite-model-find input.smt2 CVC4 version/commit: CVC4 1.8 Operating system: CentOS Linux 7