<?xml version="1.0" encoding="UTF-8"?>
<doi_batch version="5.3.1" xmlns="http://www.crossref.org/schema/5.3.1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:jats="http://www.ncbi.nlm.nih.gov/JATS1" xmlns:ai="http://www.crossref.org/AccessIndicators.xsd" xsi:schemaLocation="http://www.crossref.org/schema/5.3.1 http://www.crossref.org/schema/deposit/crossref5.3.1.xsd">
 <head>
  <doi_batch_id>aspg-40-4462-1792464976</doi_batch_id>
  <timestamp>20261020025616</timestamp>
  <depositor>
   <depositor_name>American Scientific Publishing Group</depositor_name>
   <email_address>admin@americaspg.com</email_address>
  </depositor>
  <registrant>American Scientific Publishing Group</registrant>
 </head>
 <body>
  <journal>
   <journal_metadata language="en">
    <full_title>Pure Mathematics for Theoretical Computer Science</full_title>
    <abbrev_title>PMTCS</abbrev_title>
    <issn media_type="electronic">2995-3162</issn>
   </journal_metadata>
   <journal_issue>
    <publication_date media_type="online">
     <year>2026</year>
    </publication_date>
    <journal_volume>
     <volume>6</volume>
    </journal_volume>
    <issue>1</issue>
   </journal_issue>
   <journal_article publication_type="full_text">
    <titles>
     <title>Spectral Moment Invariants and the Weisfeiler–Leman Hierarchy: Separating Power, a Fundamental Limitation, and Three Open Problems</title>
    </titles>
    <contributors>
     <person_name sequence="first" contributor_role="author">
      <given_name>Murat</given_name>
      <surname>Ozcek</surname>
      <affiliations>
       <institution>
        <institution_name>Department of Mathematics, Gaziantep University, Gaziantep, Turkey</institution_name>
       </institution>
      </affiliations>
     </person_name>
    </contributors>
    <jats:abstract>
     <jats:p>We study spectral moment invariants Σp(G) = (trAG, . . . , trApG), built from the adjacency spectrum of a graph G, as graph isomorphism invariants positioned relative to the Weisfeiler–Leman (WL) hierarchy used to characterize graph neural network (GNN) expressivity. We give three fully verified case studies. First, C6 vs. 2K3: a 1- WL-indistinguishable pair separated by Σ3 but not Σ2 or Σ4. Second, K3,3 vs. the triangular prism: a second 1-WL-indistinguishable, cubic pair, also separated at order 3, but where the fourth moment separates as well, showing separating power is graph-family dependent even at fixed order. Both examples are grounded in a general fourth-moment identity (Proposition 2.8) and a homomorphism-counting identity (Proposition 2.5) that we prove from first principles. Third, and in the opposite direction, we prove that spectral moment invariants of every finite order fail on the classical cospectral, non-isomorphic strongly regular pair srg(16,6,2,2) (the Shrikhande graph and the 4×4 rook’s graph), verified numerically to fourth order by two independent methods. We situate both directions within published enumeration data and the Godsil–McKay switching construction, compare the computational complexity of spectral-moment, WL, and general isomorphism testing, connect the limitation to Laplacian eigenvector positional encodings in Graph Transformers, and pose three precise open problems.</jats:p>
    </jats:abstract>
    <publication_date media_type="online">
     <year>2026</year>
    </publication_date>
    <pages>
     <first_page>2 2</first_page>
     <last_page>26</last_page>
    </pages>
    <publisher_item>
     <item_number item_number_type="article-number">4462</item_number>
    </publisher_item>
    <ai:program name="AccessIndicators">
     <ai:license_ref applies_to="vor">https://creativecommons.org/licenses/by/4.0/</ai:license_ref>
    </ai:program>
    <doi_data>
     <doi>10.54216/PMTCS.060103</doi>
     <resource>https://www.americaspg.com/journal/40/article/4462</resource>
     <collection property="text-mining">
      <item>
       <resource mime_type="application/pdf">https://www.americaspg.com/storage/articles/manuscripts/01KYFJ23VN8DSK7N1B6721HZY9.pdf</resource>
      </item>
     </collection>
    </doi_data>
   </journal_article>
  </journal>
 </body>
</doi_batch>
