PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 6, 20260 citationsOpen Access

Computational Invariance, Effective Witnesses, and a Structural Reduction of the P vs. NP Problem

View Full Paper
MAMichael Arias

Key Points

  • The study aims to establish a structural principle for deterministic polynomial-time computation through computational invariance.
  • Proves the Computational Invariance Theorem based on invariance of computational states.
  • Introduces concepts like bounded exposure and effective witnesses.
  • Examines the relationship between invariance and strong semantic determinism.
  • Establishes that elimination of global compatibility necessitates an explicit certificate.
  • Characterizes Globally Non–Local Irreducible Dependencies as the main obstruction to invariance.
  • Reduces the P vs. NP problem to examining these dependencies in NP-complete problems.

Abstract

This paper establishes a structural principle governing deterministic polynomial-time computation based on invariance of computational states. Building on earlier work introducing bounded exposure and effective witnesses, it proves the Computational Invariance Theorem, which asserts that no computation can eliminate global compatibility without producing a finite, explicit, and extractable certificate. The theorem is shown to be equivalent to bounded exposure normal forms and strong semantic determinism, and implies the Strong Bounded Exposure with Effective Witnesses principle. The paper identifies Globally Non–Local Irreducible Dependencies as the unique obstruction to invariance and provides a topological characterization of this phenomenon. As a consequence, the P versus NP problem is reduced to the existence of such dependencies in NP-complete problems.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Michael Arias (2026) studied this question.

synapsesocial.com/papers/698585bd8f7c464f230094eahttps://doi.org/10.5281/zenodo.18483136
Ask AI
Helpful
Bookmark
Share
View Full Paper