The <_:bnode> syntax is non-standard and is not supported through the
Jena interface of Stardog. The main reason for this is when you call
`Query.create(String).toString()` in Jena, the special bnode in the
query string turns into a regular bnode in the output string. There
are various restrictions for regular bnodes in SPARQL which causes the
errors you see.
The workaround would be to use a variable in the query and then a
filter like this:
FILTER (STR(?s) = "bnode__...")
Note that, there is no preceding "_:" when STR is applied to a bnode in Stardog.
Best,
Evren