You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Now that egglog's boolean EqSat(Theory, GroundTerm, GroundTerm) is exposed as a CQL prover, to lift it to full theorem proving it is necessary to choose between
boolean Eq(Theory t, Term t1, Term t2)
t1' = herbrandize(t1)
t2' = herbrandize(t2)
return EqSat(t U vars(t1) U vars(t2), t1', t2');
and
Eq(Theory t, Term t1, Term t2)
t1' = herbrandize(t1)
t2' = herbrandize(t2)
return EqSat(t U vars(t1) U vars(t2) U {true,false,eq,eq(x,x)=false,eq(t1',t2')=true}, true, false);
or something else entirely, to lift the current egglog prover. The second method is from unfailing knuth-bendix completion, and the first is "naive". egraphs-good/egglog#519
The text was updated successfully, but these errors were encountered:
Now that egglog's
boolean EqSat(Theory, GroundTerm, GroundTerm)
is exposed as a CQL prover, to lift it to full theorem proving it is necessary to choose betweenand
or something else entirely, to lift the current egglog prover. The second method is from unfailing knuth-bendix completion, and the first is "naive". egraphs-good/egglog#519
The text was updated successfully, but these errors were encountered: