# Faster than linear set intersection

Is there a probabilistic data structure, such as the bloom filter, for which we could perform intersection superlinearly (at the cost of false positives)? (I only know of a O(n) time complexity set intersection algorithm)