I acknowledge it may be difficult to determine the cutoff. But I'm looking for a paper clearly in one of the top twenty physics journals.

This is specifically about physics not data science or information theory.

Before 2030