The extremal inverse sum indegree energy conjecture for trees
Let be a tree on vertices. Write for its inverse sum indegree energy, and let and denote the star and path on vertices, respectively. Extremal inverse sum indegree energy conjecture. Among all -vertex trees, the tree with minimal ISI energy is , and the tree with maximal ISI energy is . The conjecture is based on numerical testing; the authors note that applying a Coulson-type integral expression has not so far yielded a proof.
References
Primary source
Sumaira Hafeez and Rashid Farooq, “Inverse sum indeg energy of graphs”, arXiv:1905.03948 (2019).
Progress summary
A reader-submitted calculation claims the conjecture is false because a seven-vertex tree has greater energy than the path, but the claim has not been independently verified.
Hafeez and Farooq proposed in 2019 that, among all trees on vertices, minimizes and maximizes inverse sum indegree energy. They described the conjecture as numerically supported, with no proof obtained from a Coulson-type integral.
Known results
- Hafeez and Farooq, 2019: the extremal statement was recorded as a conjecture; the cited work reports no proof or counterexample.
August 25, 2026 community counterexample
A submitted calculation argues that a specific seven-vertex tree satisfies , thereby disproving the path-maximization claim, and further claims counterexamples at arbitrarily large orders. The submission is unverified.
Current status (as of August 2026): The 2019 conjecture remains unproved in the published record, while an unverified community submission claims it is false for and infinitely many larger orders.
Sources
- arxiv.org
- arxiv.org
- match.pmf.kg.ac.rs
- combinatorics.org
- sciopen.com
- deepmind.google
- aimsciences.org
- scientificamerican.com
- aimspress.com
- arxiv.org
- arxiv.org
- arxiv.org
- ar5iv.labs.arxiv.org
- arxiv.org
- mathstodon.xyz
- mathstodon.xyz
- mathstodon.xyz
- quantamagazine.org
- deepmind.google
- alignment.anthropic.com
- quantamagazine.org
- www-cdn.anthropic.com
Solutions 1
CounterexampleThis solution needs a summarySee full solution
Conjecture 1 of Hafeez and Farooq, Inverse Sum Indeg Energy of Graphs (IEEE Access 7 (2019), 100860–100866; arXiv:1905.03948), asserts that among all trees on vertices the path maximizes inverse-sum-indegree energy. We disprove this at and construct infinitely many counterexamples.
For a graph , write for vertex degrees and define its symmetric inverse-sum-indegree matrix and energy by
Let consist of a central vertex joined to three internally disjoint paths of length two. Its three center-to-intermediate edges have weight , and its three intermediate-to-leaf edges have weight . Direct decomposition into symmetric and antisymmetric arm subspaces gives
For the seven-vertex path, the two end edges have weight and all other edges have weight . Its tridiagonal determinant is
Adding its positive eigenvalues and combining the two quadratic radical terms gives
All comparisons can be certified rationally. Squaring positive quantities gives
Consequently,
We next prove that failures occur at arbitrarily large orders. For , construct a tree from a -vertex spine by attaching one length-two pendant arm at every spine vertex and one additional length-two arm at each end of the spine. When , the two ends coincide, so the single spine vertex receives three arms. Every spine vertex has degree , and
Delete the four vertices of the two additional end arms, taking the principal submatrix rather than recomputing degrees. After grouping spine, intermediate, and leaf vertices, the resulting matrix is
Trace-norm contraction under orthogonal compression implies
Diagonalizing by the discrete sine transform reduces to the orthogonal direct sum of
For , the unique negative eigenvalue of is , where is the unique positive root of
Substituting yields
If , use to see that the last expression is negative. If , put ; it becomes
Since has exactly one positive root, this proves
The trace of is , so its energy is . Also is orthogonally similar to . Therefore
For even , the elementary finite cosine sum and give
Put , which is even. The inverse-sum-indegree path matrix factors as
Since , trace-norm contraction and the exact ordinary-path energy formula show that
For the last inequality, set and use .
Finally, the classical rational bounds give
Thus the path-maximality assertion fails already on seven vertices and for infinitely many trees of maximum degree three. The separate assertion that the star minimizes inverse-sum-indegree energy is not addressed here.