Toward a Tractability Frontier for Exact Relevance Certification

2026-04-08Computational Complexity

Computational ComplexityArtificial IntelligenceLogic in Computer Science
AI summary

The authors study a problem about identifying which parts of a complex system are really needed to make the best decision. They show that certain mathematical techniques can't perfectly classify or simplify this problem because of inherent complexities and exceptions. Their proof involves constructing specific counterexamples that break possible easy solutions. This means any method trying to characterize these decision problems exactly will face fundamental limits.

exact relevance certificationcoordinate-structured decision problemoptimizer-quotient realizabilitystructural predicatesclosure lawsmeta-impossibility theoremquotient entropyaffine witnessestractabilityobstruction families
Authors
Tristan Simas
Abstract
Exact relevance certification asks which coordinates are necessary to determine the optimal action in a coordinate-structured decision problem. The tractable families treated here admit a finite primitive basis, but optimizer-quotient realizability is maximal, so quotient shape alone cannot characterize the frontier. We prove a meta-impossibility theorem for efficiently checkable structural predicates invariant under the theorem-forced closure laws of exact certification. Structural convergence with zero-distortion summaries, quotient entropy bounds, and support-counting arguments explains why those closure laws are canonical. We establish the theorem by constructing same-orbit disagreements for four obstruction families, namely dominant-pair concentration, margin masking, ghost-action concentration, and additive/statewise offset concentration, using action-independent, pair-targeted affine witnesses. Consequently no correct tractability classifier on a closure-closed domain yields an exact characterization over these families. Here closure-orbit agreement is forced by correctness rather than assumed as an invariance axiom. The result therefore applies to correct classifiers on closure-closed domains, not only to classifiers presented through a designated admissibility package.