arXiv++ Combinatorics

Does computing a composition of total Boolean functions require the product of the randomized query complexities of its factors? We give a negative answer. There are families of total functions $f,g$ such that $R(f\circ g) =Θ\!\left(\frac1{\log R(g)}\right)R(f)R(g)$ This result requires first separating linearized query complexity $\LR$, as well as noisy randomized query complexity $noisyR$, from standard randomized query complexity $R$. Both constructions essentially come from totalizing existing partial function separations, albeit via different techniques. By adapting these techniques, we also construct total functions separating every adjacent pair in the chain from sabotage $RS$, max-conflict $\barχ$, and linearised query complexity $LR$. We show that AND-of-ORs separates conflict $χ$ from max-conflict $\barχ$ complexity and makes conflict and sabotage incomparable. Finally, randomized complexity can be smaller than its composition limit $R^*$ by the maximal factor $Θ(\log R)$. The total--total composition separation is also formalized in Lean.

We give a new proof that 1-extension preserves generic local rigidity of bar-joint frameworks in all dimensions. The proof does not use a special position argument, which makes the approach potentially useful for X-replacement in 3-space. Instead, the balancing conditions recently introduced by the first author use directed paths in the line graph, and by modifying the end of these paths, we transfer a self-stress on the 1-extended graph back to the original. Treating the new vertex as variable allows us to extract the balancing condition of the smaller graph as one coefficient of a homogeneous quadratic. A Lean 4 formalization of the main results accompanies the paper.

A single-track Gray code of length $n$ is a cyclic Gray code of $n$-bit words in which every column is a cyclic shift of the first. For prime $n$ the longest possible period is $2^n-2$. Whether it is attained for every prime has been open since Hiltgen, Paterson and Brandestini introduced these codes in 1996, and optimal codes were known for primes $n\le 19$. We show that a conjecture of Degni and Drisko (2007) on the reflected Gray code order of necklaces implies optimal single-track Gray codes for every prime. Their list of necklaces, conjecturally a Gray path for prime $n$, closes into a Hamilton cycle of the necklace graph after $(n-3)/2$ explicit Pósa rotations, and such a cycle lifts to an optimal code. This implication is formalized in Lean 4. Together with computation it gives optimal single-track Gray codes for every odd prime $n\le 47$: for $n\le 37$ explicit codes checked word by word by a program independent of the generator, and for $n=41$, $43$ and $47$ codes obtained from the implication after the Degni-Drisko property was verified at these lengths by two independently written programs.

We prove that $Z(P(n,3))=8$ for every $n\ge13$ and $Z(P(n,4))=10$ for every $n\ge18$, with both stabilization thresholds sharp, and give the exact values for all smaller admissible $n$. The first equality proves Krishnan's Conjecture 5; his Proposition 4 supplies its upper bound. The lower bounds use the external boundary of the first forcing sources: ten vertices have boundary at least eight in $P(n,3)$ for $n\ge14$, and sixteen vertices have boundary at least ten in $P(n,4)$ for $n\ge20$. Deletion within an empty column interval reduces these assertions to finite checks, and forts settle the remaining cases. The $k=3$ stabilization theorem has a Lean 4 formalization; the $k=4$ proof uses 89 boundary formulas and 11 fort formulas with separately checked DRAT certificates, together with explicit forcing sequences.

String graphs are intersection graphs of continuous curves in the plane. In 2010, Fox and Pach proved that string graphs with $m$ edges admit balanced vertex separators of size $\mathcal{O}(m^{3/4}\sqrt{\log m})$ and conjectured that this bound can be improved to $\mathcal{O}(\sqrt{m})$, which would be optimal. After partial progress, Lee claimed in 2017 to prove this conjecture, and more generally to prove analogous separator theorems for region intersection graphs over minor-closed graph families. A key intermediate step in Lee's approach was a padded decomposition scheme for region intersection graphs. In 2024, Korhonen and Lokshtanov extended Lee's approach to the broader setting of graph classes excluding a fixed graph as an induced minor. These results have since found applications in a wide range of works. Later in 2024, a flaw was discovered in the construction of Lee's padded decomposition scheme, invalidating Lee's as well as Korhonen and Lokshtanov's separator results, and many works relying on them. In this work, we prove that string graphs with $m$ edges admit balanced separators of size $\mathcal{O}(\sqrt{m})$, and more generally obtain analogous $\mathcal{O}(\sqrt{m})$ bounds for region intersection graphs over fixed minor-closed graph families and for graphs excluding a fixed graph as an induced minor, thereby repairing the vast majority of the affected works. Furthermore, our balanced separators can be found in polynomial time. Our proof uses a different approach from Lee's and does not rely on padded decompositions. Since padded decompositions have also found applications independent of the separator theorems, particularly for string graphs, we also prove a padded decomposition result for region intersection graphs that is optimal for string graphs. We have formalized both the balanced separator and the padded decomposition results in Lean 4 with mathlib.

Two players alternately claim cells of an $n \times n$ grid; the first to hold all four corners of an axis-parallel square wins, and a full board without one is a draw. The first player has a winning strategy if and only if $n \ge 5$, which answers a question of Erickson. For $n \ge 6$ one strategy serves every board: a rule with no lookup table whose moves after the second reply are all forcing, built around the plus, a stone with the four cells at a common distance from it along the axes, a shape proposed for this game by Sirgedas in 2010. The rule is checked on the boards of size $6$ to $17$, and a simulation theorem carries one checked run on the $23 \times 23$ board to every larger board. The proofs are formalized in Lean 4, and the theorem for the rule states that the tree of its own moves is winning.

Razborov's flag algebra method is one of the most successful tools for proving asymptotic density inequalities in extremal graph theory. It reduces an inequality about all large graphs avoiding a forbidden pattern to finitely many algebraic conditions, and a certificate satisfying them can be found by semidefinite programming. We formalize the method for finite simple graphs in Lean 4, building on Mathlib. The formalization covers flags and their densities, the flag algebra as a quotient vector space, positive homomorphisms as limits of graph sequences, Razborov's random extensions, and the downward operator that averages out labels. On top of this theory, we build verified computation of flag densities and a tactic that turns certificates found by semidefinite programming into Lean proofs. The certificate serves only as candidate data; Lean checks every fact the proof uses. With this development we verify seven Turán-type upper bounds. Adding lower-bound constructions proved in Lean yields the asymptotic form of Mantel's theorem and the asymptotic Erdős pentagon theorem $π(C_5;K_3)=24/625$; to our knowledge, the latter has not previously been formalized in a proof assistant. We also prove two inequalities of Goodman. All density theorems reported here depend only on the standard axioms of Lean's logic. Formalization also led us to treat forbidden subgraphs with a different semantics. Instead of building a separate flag algebra for each forbidden family, we keep a single algebra of all graphs and impose the constraint only when interpreting an inequality, through randomly labeled limits of graphs satisfying it. We prove that this ensemble semantics yields the intended density bounds, and that a condition we call root-plantability characterizes when it agrees with the standard semantics.

Let $f(n)$ be the largest integer such that every set of $n$ points in the plane contains $f(n)$ points no two at distance $1$, and let $m_1$ be the supremum of the upper densities of Lebesgue measurable planar sets with no two points at distance $1$. Larman and Rogers observed that $f(n) \ge m_1 n$. We prove that $f(n)/n \to m_1$; that is, $f(n) = (m_1 + o(1))n$. This is the first half of a conjecture of Dúcz and Varga, and with the bound $m_1 \le 0.247$ of Ambrus, Csiszárik, Matolcsi, Varga and Zsámboki it gives $f(n) \le (0.247 + o(1))n$. The proof builds a random independent subset of the algebraic plane, invariant under algebraic translations and rotations, whose probability of containing the origin is at least the worst finite independence ratio, and turns it into a measurable planar set of at least that upper density. Its only non-classical input is the Haar rigidity theorem for wild character laws from the OpenAI preprint The Euclidean plane is not five-colorable, which has a machine-checked Lean proof. The whole argument has been formalized in Lean 4 on top of it.

We exhibit a fixed $4\times 4$ binary matrix with no polynomial induced ordered removal bound. At depth $h\geq 1$, let $m=2^h$. Our explicit host has order $(14h+2)m$, exactly $2m^6+\frac{1}{2}(m-2)m^5$ induced copies, and exact editing distance $m^2$ cells. Its normalized editing distance is $Θ(h^{-2})$, while its induced-copy density is $Θ(4^{-h}h^{-8})$. The lower bound holds against arbitrary cell edits through an injective family of $m^2$ protected selections. An exhaustive arithmetic classification identifies every induced copy and proves that a matching canonical repair creates none. The original-copy transversal number is exactly $m$, yielding an unbounded gap between hitting existing copies and eliminating all copies. The identities and the obstruction at every depth are formalized in Lean, with verification sources included as ancillary files.

Let $G$ be a finite bipartite graph and let $\widetilde G$ be an $M$-fold cover. We compare independent sets of $\widetilde G$ with those of the trivial cover, consisting of $M$ disjoint copies of $G$, after prescribing how many vertices are selected above each vertex of $G$. A conjecture from 2011, motivated by the Bethe approximation, asserts that the trivial cover maximizes every such count. We prove this for every feasible occupation profile whose active graph has treewidth at most two or is chordal bipartite. Equality holds exactly when the restricted cover is flat, meaning that fibre relabelings make it trivial. For $M\ge2$ and positive feasible occupations, a nonflat cover of a connected bipartite graph of treewidth at most two has a deficit of at least twice the coefficient of $p_G^{M-2}$ with every occupation reduced by one. A holonomy-dependent bound improves the factor to three for odd-order holonomy; both uniform constants are sharp. Comparison and rigidity survive finite compositions along an edge or a vertex, and quantitative bounds from one piece propagate to the whole graph, giving results beyond the two original classes. On any bipartite base, the first nonzero homogeneous defect is determined by the shortest cycles with nontrivial holonomy. At singleton occupations, our counts are numbers of DP-colorings of full covers. The dual DP color function on these classes and their finite seam compositions equals the chromatic polynomial, with equality attained precisely by canonical labelings. Dong's chromatic-polynomial inequality then yields the Dual-DP Shameful inequality on the same graph class. The main theorems have been formalized in Lean 4.

Under the conditional framework breaking the longstanding O(n log n) integer multiplication barrier (originating from OpenAI Preprint #109), the asymptotic time complexity exponent T(n) = O(n (log n)^{1-κ}) is governed by bounding the second moment E[X^2] < 1 of recursive branching processes over trilinear tensor carrier graphs. In this work, we present a constructive combinatorial optimization over the alternating carrier bipartite matching graphs of the Schönhage trilinear tensor decomposition at dimensions h=23 and h=25. Through an alternating augmenting path search, we uncover a non-trivial cluster-unlocking cascade that eliminates 185 carrier roles (95 roles in h=23 and 90 roles in h=25), reducing the global physical wire width by 377,890 wires from the reviewed baseline W = 177,284,805 down to W = 176,906,915. Evaluated over the exact 10^{-14} rational grid, this wire reduction certifies an improved second-moment coefficient AB = 5,164,059 / 125,000,000,000 and establishes the sharpened conditional exponent κ= 826,215,307 / 20,000,000,000,000 \approx 4.131076535 \times 10^{-5}, strictly surpassing the preceding record on this architecture (κ\approx 4.123864 \times 10^{-5}). All 47 polynomial constraints and 7 rational security margins are audited in exact Q-arithmetic with zero floating-point error, and the fundamental wire reduction identity is formally verified in the Lean 4 interactive theorem prover without unproven axioms.

Random spanning forests provide probability models for partitions and Monte Carlo methods on graphs. In these applications, the distribution produced by a sampler determines which forest functionals can be estimated. The Aldous--Broder algorithm constructs a spanning tree from the first-entrance edges of a covering random walk. Its distribution is known, but a direct calculation requires summing over all orders in which a fixed tree can be discovered. We evaluate this sum through an identity for rooted forests and real matrices with zero row sums. Under a nonsingularity condition on the principal submatrices encountered during growth, the sum of products of harmonic extension entries equals a principal cofactor of a matrix on the roots. With one root, the sum is one. The identity gives a direct proof of the extended Aldous--Broder theorem for every finite irreducible Markov chain started at a fixed state, without assuming reversibility or bidirectional transition support. It also gives the conditional distribution of the discovery order and the law of the forest obtained by stopping after all nonroot vertices have been visited. A completion weight describes the difference between this stopped-walk law and the weight-proportional forest law, including the support restriction needed for importance correction. Numerical examples illustrate the sampling implications. The identity, the Aldous--Broder theorem, and the stopped-forest law are formalized in Lean4.

Csernák and Soukup introduced the stacking number of a graph: the least integer $t\ge2$ such that every configuration of $t$ pebbles can be transformed by pebbling moves into a configuration supported on one vertex. They conjectured an explicit formula for the stacking number of a tree in terms of the degrees of its vertices and their distances from a chosen root. We prove this conjecture for every finite tree with at least two vertices. We also show that whether a configuration on a tree can be stacked at a given vertex is decided exactly by the sign of an integer score built from recursive branch messages; consequently, stackability can be tested with a linear number of arithmetic operations. The proof combines this criterion with an explicit zero-score obstruction and a weighted cancellation argument for arbitrary nonstackable configurations. The main theorem and the stackability criterion are formalized in Lean 4, and the main theorem has passed Palomar mechanical verification; it is publicly registered as PALOMAR-2026-09-25-000010.

Let $k$ be a field of characteristic $p\ge0$. We prove that $m$ points and $n$ lines in $k^2$ determine at most $3(mn)^{2/3}+m+n+2mn/p$ incidences, the last term being omitted in characteristic zero. Over the prime field $\mathbb{F}_p$ the coefficient of $mn/p$ can be replaced by $1$. The proof uses the polynomial method, and for $m=n$ the bound is sharp up to an absolute constant over prime fields. As applications, over prime fields in which $-1$ is not a square we obtain the $L^2\to L^r$ extension estimate for the paraboloid in $\mathbb{F}_p^3$ for $r>10/3$. Over every odd prime field, we show that a two-source extractor construction of Bourgain has exponentially small error at every min-entropy rate greater than $1/3$. We also improve sum-product estimates for small sets in positive characteristic and obtain projection and Furstenberg estimates over prime fields. The incidence inequalities with exact constants have been formalized in Lean.

We prove the conjecture of Huang, Jiang, and Oblomkov (HJO) giving a geometric extension of the Rogers--Ramanujan and Andrews--Gordon identities for every torus-knot singularity $X^a=Y^b$ with coprime $1<a<b.$ For a prime power $q$, let $\mathcal{NC}_n^{a,b}(\mathbb F_q)$ denote the set of pairs of commuting nilpotent $n\times n$ matrices $(A,B)$ over $\mathbb F_q$ satisfying $A^a=B^b$. We establish the threefold equality between their normalized counts, the HJO $q$-series $Z_{a,b}$, and the explicit infinite product $P_{a,b}$: \[ \underbrace{\vphantom{\Bigg|} \prod_{m\geq1}(1-q^{-m}) \Biggl(\sum_{n=0}^{\infty} \frac{\lvert\mathcal{NC}_n^{a,b}(\mathbb F_q)\rvert} {\lvert\operatorname{GL}_n(\mathbb F_q)\rvert}\Biggr) }_{\text{point count}} = \underbrace{\vphantom{\Bigg|}Z_{a,b}(q^{-1}) }_{\text{\(q\)-series}} = \underbrace{\vphantom{\Bigg|}P_{a,b}(q^{-1}) }_{\text{infinite product}}. \] Our main result is a stronger finite identity: the rank $N$ HJO sum equals $(q;q)_N$ times the generating function for balanced cylindric partitions with entries bounded by $N$. Taking $N\to\infty$ yields the HJO conjecture. The proof combines the compositional rational shuffle theorem of Bergeron--Garsia--Leven--Xin and Mellit with a multiplicativity theorem for slope operators and a determinantal model for bounded cylindric partitions, linked by a common $q$-difference equation. The finite identity and the HJO conjecture have been formalized unconditionally in Lean by AxiomProver.

Harwit and Sloane conjectured that every nonsingular entrywise-nonnegative matrix $A\in\mathbb R^{n\times n}$ satisfies $\|A^{-1}\|_F\ge 2n(n+1)^{-1}\|A\|_{\max}^{-1}$, with equality precisely for positive multiples of $S$-matrices. Zhang has given a complete proof of this conjecture by a centered pseudoinverse and spectral variance method. We present an independently obtained, structurally different proof of the strict even-dimensional inequality. Starting from the structural identities of Frankel and Urschel, we derive an exact global defect budget and combine binary rounding, fixed intersections, and Gram projection. A ten-row obstruction handles every even $n\ge66$; a finite exact calculation handles $4\le n\le64$, $n\ne6$; and a multi-column energy argument treats $n=6$. The order-two case is elementary. The even-dimensional argument is formalized in Lean 4, conditional on Frankel--Urschel Lemma 2.1 as an explicit external mathematical input. The finite evaluations use Lean's native evaluator; their trust boundary and exact certificates are documented. Together with Cheng's odd-dimensional theorem, the argument recovers the full S-matrix theorem and its equality characterization.

For every integer $k\ge1$, we prove that each nonempty bounded set in real $\ell_1^{4k+1}$ has an ambient enclosing ball of radius at most $(4k-1)/(4k)$ times its diameter. The bound is sharp in dimension five: $J(\ell_1^5)=3/4$. If a Hadamard matrix of order $4k$ exists, then $J(\ell_1^{4k-1})=J(\ell_1^{4k})=J(\ell_1^{4k+1})=(4k-1)/(4k)$. The upper bound for every $k\ge1$ has been formalized in Lean.

We prove Sabidussi's compatibility conjecture. Let $G$ be a finite connected multigraph in which every vertex has even degree and the minimum degree is at least four, and let $T$ be an Euler tour of $G$. The edges of $G$ can be partitioned into circuits (connected $2$-regular subgraphs) so that no circuit contains two edges used consecutively anywhere in $T$. In fact, the edges can be four-coloured so that every such pair receives different colours and every colour class has even degree at every vertex. We use a counting argument based on the Chevalley-Warning theorem to show that a four-colouring with the required properties exists. Splitting each colour class into circuits then gives the desired compatible decomposition. Formalization in Lean 4 is also available in the author's github.

In 1965, D. H. Lehmer conjectured that the permutations of every multiset admit an imperfect Hamiltonian traversal by adjacent swaps: a walk in the neighbor-swap graph that visits every word, with some words visited twice in order to reach a neighbor and return. The question is posed as an unsolved research problem in Knuth's Art of Computer Programming. Verhoeff (2017) chose the stutter words, in which every domino is a double, as the words to be reached this way, and reformulated the conjecture as the Hamiltonicity of the graph $N(S)$ on the non-stutter words, with two exceptional families --- binary signatures with an odd multiplicity, and the permutations of $(2k,1,1)$ --- that admit a Hamiltonian path but no cycle. This article proves the reformulated conjecture, and with it Lehmer's conjecture. The key structure is a partition of the words into hypercubes: the swaps inside dominoes turn each class of words with the same domino contents into a hypercube, and the stutters are exactly the $0$-dimensional classes. When every multiplicity is even, Hamiltonian cycles of the hypercubes are glued along a spanning tree, with no finite check. The case of exactly one odd multiplicity reduces to the all-even case and to a theorem of Stachowiak (1992), the one inherited Hamiltonicity input, which also settles two or more odd multiplicities. The only finite ingredients are two explicit cycles, of 28 and 84 words. Every construction is implemented in Python and checked against brute-force graphs, and the proof is formalized in Lean 4 over Mathlib.

For a graph $H$, let $f(n,e,H)$ be the least number of colors in an edge-coloring of some $n$-vertex graph with at least $e$ edges in which every copy of $H$ is rainbow. Burr, Erdős, Graham, and Sós conjectured that $f(n,\lfloor n^2/4\rfloor+1,C_{2k+1})=(1/8+o(1))n^2$ for every fixed $k\ge3$, and Bucić, Chen, and Ma recently proved this for all $k\ge4$. We prove the remaining case $k=3$: \[ f\left(n,\left\lfloor n^2/4\right\rfloor+1,C_7\right) =\left(\frac18+o(1)\right)n^2. \] The lower bound rests on a weighted palette inequality, which we prove with an exact rational certificate on five sampled vertices. Its main ingredients are a fractional matching of compatible triangular edges and private resources attached to nontriangular edges. A stable form of the inequality, combined with regularity, triangle removal, and a direct argument for graphs close to bipartite, transfers the bound to arbitrary edge-colorings. We also describe a Lean 4 formalization of the conjecture for every fixed $k\ge3$, which combines the new seven-cycle proof with a formalization of the Bucić-Chen-Ma argument for $k\ge4$.