🤖 AI Summary
This work investigates the intrinsic algebraic complexity cost of Boolean summation (∑_{w∈{0,1}^n})—a fundamental operation whose efficient elimination remains an open challenge in algebraic complexity theory.
Method: We systematically quantify lower bounds on algebraic elimination of single-bit Boolean summation, combining arithmetic circuit lower-bound techniques, polynomial-family reductions, and explicit constructions of combinatorial identities.
Contribution/Results: We establish, for the first time, the essential structural role of summation within the VP/VNP framework. As a direct consequence, we obtain the simplest known proof of the VNP-completeness of the permanent—relying solely on the irreducibility of basic Boolean summation, without invoking determinant simulation or deep encoding machinery. This result strengthens the fundamental separation between the algebraic computational power of the permanent and the determinant, and provides a crucial algebraic bridge linking the P vs NP and VP vs VNP questions.
📝 Abstract
The P versus NP problem is about the computational power of an existential $exists_{w in {0,1}^n}$ quantifier. The VP versus VNP problem is about the power of a boolean sum $sum_{w in {0,1}^n}$ operation. We study the power of a single boolean sum $sum_{w in {0,1}}$, and prove that in some cases the cost of eliminating this sum is large. This identifies a fundamental difference between the permanent and the determinant. This investigation also leads to the simplest proof we are aware of that the permanent is VNP-complete.