I was testing out all the possible ways to improve the performance of hashing and matching for the Ligity paper. Here is one of the cases where I benchmarked a bloom filter: