REVIEW 5 minor 15 references
The semi-inducibility of the blue--blue--red path on four vertices
T0 review · 0 major / 5 minor · reviewed 2026-07-31 · deepseek-v4-flash
Pith's one-line read The paper determines the exact maximum limiting density of semi-induced blue–blue–red paths on four vertices, resolving the last open four-vertex case.
desk verdict This paper really does settle the last open four-vertex semi-inducibility case, and the proof, though long and intricate, appears sound. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central machinery is the degree quotient combined with a degree-square tie-break. The functional Φ(W) measures weighted H3 density and can be written as β(1-β) - ∫ψ(d) - E_W(d) in terms of the red degree function d. For any quotient W^P that merges degree information, Lemma 4.3 shows that a nontrivial quotient strictly decreases Φ whenever the chosen maximizer also minimizes D(W) = ||d||²₂; equality would produce a second maximizer with smaller D. This strictness forces the degree distribution of an extremal graphon into a single high-degree atom A of weight > 1/2 that is a clique, with a low-degree remainder, collapsing the problem to the two-variable family F(a,x) and its convex-envelo
What would settle it
Evaluate the normalized semi-induced H3 count of the proposed construction for large n—a clique of size a*n disjoint from a regular graph of degree x*n—and compare with p*; any asymptotic excess would refute the upper bound. A direct computer search over small step graphons for Φ > p* would also disprove the theorem.
Extended reading notes
Core claim
Theorem 1.1 states that I(H3) = p*, where p* = F(a*, x*), x* is the unique root of 32x^3 - 40x^2 + 13x - 1 = 0 in (0, 1/4), and a* = (4x*^2 - 7x* + 2)/(3 - 8x*). Equality is attained by graphs of the form K_{(a*+o(1))n} ⊔ R_n, where R_n is asymptotically regular with every vertex of degree (x*+o(1))n, equivalently relative degree λ*+o(1) inside R_n. The proof works with graphon limits and shows that a maximizer chosen to minimize the degree-square functional must have a clique A of weight > 1/2, no edges between A and its complement, and degree at most 1/2 on the complement. This split structure turns the upper bound into the two-variable inequality F0(a,x) ≤ p*, proved by calculus and conve
Load-bearing premise
The load-bearing premise is Lemma 4.3: for a chosen maximizer that minimizes the degree-square functional, every nontrivial degree quotient strictly decreases the objective; if equality ever occurred, the proof could not force the clique-plus-low-degree split.
Editorial extensions
If this is right
- The exceptional four-vertex red-blue graph is now settled, completing the classification of semi-inducibility for all non-complete red-blue graphs on four vertices.
- The tie-break argument produces an extremal graphon of a forced split form: a clique of weight > 1/2 plus a remainder whose degrees are at most 1/2.
- The upper-bound proof reduces to the explicit inequality F0(a,x) ≤ p*, so any graph's H3-density is controlled by a two-parameter degree summary.
- The lower-bound construction is simple and explicit at every n—a clique and a cyclic regular graph—showing the bound is asymptotically sharp.
Reading between the lines
- The degree-square tie-breaking device is likely transferable to other semi-inducibility problems: it converts an infinite-dimensional graphon maximization into a one-dimensional degree analysis whenever degree quotients preserve the objective and the tie-break forces strictness.
- One can probe the same problem with a prescribed red-edge density by maximizing F(a,x) under the constraint β(a,x) = c; the explicit formulas here suggest a piecewise algebraic density profile with a phase change near the boundary of the split-regular region.
- A computer search over small step graphons could independently test p*: a step graphon with Φ > p* would expose a gap in the quotient argument, while agreement would corroborate the exact value.
- The structural dichotomy—either a majority clique atom exists or the density is bounded below 3/20—may generalize to other red-blue trees with a blue prefix path, since the same half-slope inequality (57) governs the high-degree rows.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper determines the exact semi-inducibility constant for the four-vertex red-blue path H3 with edge colors blue–blue–red. It proves that I(H3)=p*, where x* is the unique root in (0,1/4) of 32x^3-40x^2+13x-1=0, a*=(4x*^2-7x*+2)/(3-8x*), and p*=F(a*,x*); equality is attained by a graph sequence of the form K_{(a*+o(1))n} ⊔ R_n with R_n asymptotically regular. The proof uses graphon compactness, a degree-square-minimizing maximizer, weighted vertex quotients, KKT-style edge perturbations, and a structural reduction to a clique joined to a low-degree part. The final upper bound is reduced to a two-variable optimization, which is solved exactly with explicit polynomial and Bernstein identities.
Significance. This resolves the last open four-vertex case from the Bodnár–Pikhurko classification of non-complete red-blue graphs, giving a new exact semi-inducibility value. The method is of independent interest: the degree-square tie-breaking among maximizers, the half-slope inequality, and the atomization argument form a coherent and largely self-contained structural toolkit. The lower-bound construction is explicit, and the upper-bound proof is independent of it; the analytic estimates are supported by concrete Bernstein expansions. I see no circularity or parameter fitting. If the proof is correct, this is a valuable contribution to the semi-inducibility literature.
minor comments (5)
- [§4.1 (notation)] The symbol A is used both for the σ-algebra σ(d) in (45) and later for the clique vertex set in Theorem 4.12 and Theorem 4.1. This overload makes the structural section harder to read; consider using a script letter or Σ for the σ-algebra.
- [Lemma 4.5] In the non-atomic part, the transition from Lebesgue–Besicovitch differentiation to the choice of radii satisfying the small L1 condition and then to (60) is compressed. A sentence stating that at a µ-Lebesgue point one first makes the normalized L1 error < η_k^2 and then applies Markov would help.
- [Lemma 4.14] The argument from (104) to (106) is terse where R_a crosses from positive to nonpositive. It is correct, but the reader must infer that when R≤0 the bound comes directly from U≤U0≤153/1024, and when R_a remains positive the maximum is V(1/2)=F_aux. Expanding this by a few lines would improve readability.
- [Lemma 4.15] The sign checks in (138) and (145)–(146) are correct, but the connection between the interval [1/2,1/√2] and the Bernstein variable t∈[0,1] is stated only implicitly; make explicit that the relevant t-range is a subinterval.
- [§4.1] The line 'Put d(x,y):=(d(x),d(y))' after (46) is unused; it can be deleted.
Circularity Check
No circularity: upper bound and lower bound are independently established.
full rationale
The derivation is self-contained. The upper bound (Theorem 1.1 via Lemma 2.3, Theorem 4.1, Lemmas 4.14/4.15, and eq. (152)) establishes Φ(W) ≤ p* for every graphon W, where p* is defined independently through the polynomial system (3)–(4). The lower bound is an explicit construction K_{a*n} ⊔ R_n evaluated at (a*, x*) (eq. (153) and Section 5), not a parameter fitted to match the upper bound. The structural reduction rests on Lemma 4.2–4.3, which derives strict decrease of Φ under nontrivial degree quotients from the D-minimizing choice of maximizer; this is an internally proved tie-breaking argument, not an imported uniqueness theorem. Lemma 4.5 derives the half-slope inequality from local quotient perturbations and Lebesgue differentiation, not from the target equality. Lemmas 3.3 and 3.4 solve the two-variable optimization independently of the structural theorem. The only self-citation, [DH26] in the introduction, is contextual and not used in any proof. No equation in the paper reduces to an assumed answer; in particular, (49)–(50) plus (53) are derived identities, not renamed predictions. No limitation or missing-support passages affecting circularity were found.
Assumptions & free parameters
assumptions (3)
- standard math Graphon compactness and cut-continuity of finite homomorphism densities.
- standard math Fubini's theorem, Jensen's inequality, and Lebesgue-Besicovitch differentiation for finite Borel measures.
- domain assumption The semi-induced density extends to graphons and the finite-graph value differs by O(1/n) (Lemma 2.3).
Cite this review
Pith. "Pith review of The semi-inducibility of the blue--blue--red path on four vertices." pith.science (2026). https://pith.science/paper/FUUMMAC4
@misc{pith2026260727911,
author = {Pith},
title = {Pith review of: The semi-inducibility of the blue--blue--red path on four vertices},
year = {2026},
howpublished = {\url{https://pith.science/paper/FUUMMAC4}},
note = {Machine review of arXiv:2607.27911}
}
abstract
For an $n$-vertex graph $G$, let $N(H_3,G)$ be the number of injective labeled copies of the red-blue path $H_3$ for which the two blue pairs are mapped to non-edges of $G$ and the red pair is mapped to an edge of $G$. We determine the maximum limiting value of $N(H_3,G)/n^4$ and give an extremal construction, which is the disjoint union of a clique and an asymptotically regular graph. The proof uses weighted vertex quotients and degree-square tie-breaking. We thereby resolve the exceptional four-vertex case left open in the recent classification of non-complete red-blue graphs.
Reference graph
Works this paper leans on
- [1]
- [2]
-
[3]
L. Bodn\'ar, J. Gao, J. Le\'on, X. Liu, O. Pikhurko, and S. Sun, The inducibility of 6-vertex graphs, arXiv:2606.00290, 2026
arXiv 2026
-
[4]
Bodn\'ar and O
L. Bodn\'ar and O. Pikhurko, Semi-inducibility of 4-vertex graphs, Discrete Appl. Math. 392 (2026), 303--323
2026
-
[5]
Bodn\'ar and O
L. Bodn\'ar and O. Pikhurko, Some exact inducibility-type results for graphs via flag algebras, Electron. J. Combin. 33 (2026), no. 2, Paper No. P2.39
2026
-
[6]
Bollob\'as, Y
B. Bollob\'as, Y. Egawa, A. Harris, and G. Jin, The maximal number of induced r -partite subgraphs, Graphs Combin. 11 (1995), 1--19
1995
-
[7]
J. I. Brown and A. Sidorenko, The inducibility of complete bipartite graphs, J. Graph Theory 18 (1994), 629--645
1994
-
[8]
H. Chen and J. A. Noel, On alternating 6-cycles in edge-coloured graphs, Electron. J. Combin., to appear; arXiv:2505.09809, 2025
arXiv 2025
Show all 15 references
-
[9]
Deng and J
J. Deng and J. Hou, Fixed-density profiles for the semi-induced 4-vertex star, arXiv:2606.23351, 2026
2026 arXiv
-
[10]
G. B. Folland, Real Analysis: Modern Techniques and Their Applications, 2nd ed., John Wiley & Sons, 1999
1999
-
[11]
Hatami, J
H. Hatami, J. Hirst, and S. Norine, The inducibility of blow-up graphs, J. Combin. Theory Ser. B 109 (2014), 196--212
2014
-
[12]
Lov\'asz, Large Networks and Graph Limits, American Mathematical Society, 2012
L. Lov\'asz, Large Networks and Graph Limits, American Mathematical Society, 2012
2012
-
[13]
Lov\'asz and B
L. Lov\'asz and B. Szegedy, Limits of dense graph sequences, J. Combin. Theory Ser. B 96 (2006), 933--957
2006
-
[14]
Pippenger and M
N. Pippenger and M. C. Golumbic, The inducibility of graphs, J. Combin. Theory Ser. B 19 (1975), 189--203
1975
-
[15]
A. A. Razborov, Flag algebras, J. Symbolic Logic 72 (2007), 1239--1282
2007
Reviewed July 31, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.