There are now many examples of "natural" (no mention of circuits) problems which have been proven to be complete for classical TFNP classes---these include Consensus Halving for PPA and Blichfeldt for PPP. Is there such an example which might be considered the simplest?
-
$\begingroup$ Simplest in what sense? Anyhow, my favorite is Necklace Splitting. $\endgroup$domotorp– domotorp2025年11月08日 13:34:03 +00:00Commented Nov 8 at 13:34
1 Answer 1
Maybe WEAK-PIGEON, which is complete for PWPP by definition? The input is a circuit $C: \{0, 1\}^n \to \{0, 1\}^{n - 1}$, and the goal is to find distinct $x, y \in \{0, 1\}^n$ such that $C(x) = C(y)$.
-
$\begingroup$ That is not a natural problem $\endgroup$Stefan G.– Stefan G.2025年11月04日 13:38:13 +00:00Commented Nov 4 at 13:38