The math that powers YouTube's 'diverse' recommendations is the same math that controls physical rockets.
April 15, 2026
Original Paper
Connections Between Determinantal Point Processes and Gramians in Control
arXiv · 2604.08913
The Takeaway
This paper bridges two seemingly unrelated fields: recommendation systems and linear dynamic control. It establishes that observability and controllability Gramians—the foundations of how we measure system state—are actually Determinantal Point Processes (DPPs). DPPs are the gold standard for creating 'diverse' subsets in machine learning. This means we can now use tools from modern data science to solve classical engineering problems in robotics and physical systems. It opens up a whole new toolbox for control theory using modern recommendation engine math.
From the abstract
Determinantal point processes (DPPs) are probability models over subsets of a ground set that favor diverse selections while suppressing redundancy. That is, they tend to assign higher likelihood to collections whose elements complement one another instead of repeating the same information. For example, in recommendation systems, a DPP prefers showing users several relevant items that differ in content or style, rather than many near-duplicates of essentially the same item. Although DPPs have be