Every total polynomial-time SAT-CNF candidate has an error input
OpenVMPvsNP.pn2_sat_no_polytime_deciderDoes every total uniformly polynomial-time Boolean-answering deterministic multitape machine have an input on which it gives the wrong SAT-CNF answer?
Let L be the shared binary SAT-CNF language, including rejection of malformed encodings. Let H(M,y,t,b) denote actual local execution halting by t transitions with Boolean answer b. For every machine M and positive integers c and k, the target is
Total bounded halting is a premise, but correctness and assumed failure are not. The error input may depend on M, c, and k; one common witness for every candidate is not required. A failure of one algorithm or a restricted model does not settle this universal target.
Formalization Note PA and PN2 import the same published definitions. This is an OPEN root statement with an explicit proof placeholder, not a root proof or an admitted P-versus-NP conclusion. The concrete representation correspondence remains subject to faithfulness review.
import Definitions.Def_VMPvsNP_SATModel
theorem VMPvsNP.pn2_sat_no_polytime_decider :
∀ M : VMPvsNP.Machine, ∀ c k : Nat,
0 < c → 0 < k →
(∀ y : List Bool, ∃ answer : Bool,
VMPvsNP.HaltsWithin M y (VMPvsNP.polynomialBound c k y) answer) →
∃ x : List Bool, ∃ answer : Bool,
VMPvsNP.HaltsWithin M x (VMPvsNP.polynomialBound c k x) answer ∧
¬ (answer = true ↔ VMPvsNP.SATCNF x) := by sorry