Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Hereditary span bound
Proved. Let and . There is no graph with
such that every satisfies
where , .
Consequently, in a hypothetical coloring counterexample, every induced color graph of independence number at most has at most vertices in this range. This is not a proof of Problem 617 for an infinite set of parameters: it has the additional small-independence hypothesis.
0. A self-contained finite partition lemma
The short induction in Section 3 of the independent input checks out. It proves the precise finite stability statement used here, so no external stability theorem is needed for the argument below.
Write . If is -free on vertices, it has a partition into possibly empty parts with at most internal edges.
Induct on , with immediate; an empty graph is also immediate. Choose a maximum-degree vertex , and put , , . The graph is -free. Its inductive partition, together with , has at most
internal edges. Maximum degree gives
Also , by taking a balanced -partite graph on and joining it completely to the independent part . Combining the two inequalities proves the claim.
If and is the complete multipartite graph of this partition, then its number of missing cross-edges is
and . This is the finite content of Füredi's Theorem 1 and Corollary 2. The supplied theorem and proof were checked; the induction above is the same maximum-degree mechanism in recursive form.
1. Quadratic-scale extraction
The span theorem and its equality case can be proved together. Set , where , and assume and all inherited caps. Put . Its Turán deficit is
For the exact difference is ; for it is . The finite partition lemma gives parts , at most internal -edges, and at most
target edges between parts. Comparing squared part sizes gives
Since , every part has order at least .
Mark vertices whose target cross-degree is at least , obtaining with . Suppose an internal -edge has two unmarked endpoints. In each other part, their common -neighborhood has at least vertices. Independent uniform choices from these neighborhoods have expected target-edge count at most
Some choice, together with , is a forbidden in . Thus are nonempty target cliques. The local cap forbids , so write
If , the overloaded case is already impossible, and the balanced case already consists of full cliques. Hence assume .
2. Assigning owners
For , put and . Applying the cap to and subtracting its clique and incident edges gives
when . When both deficits vanish, the stronger bound gives the same final inequality. The relevant set sizes lie between and since .
Set . Every exception has an owner with . Otherwise random representatives from all its nonneighbor lists have expected target-edge count less than
For , setting bounds this by ; for the first term is . An independent choice together with contradicts .
3. Weighted completion without a concentration loss
For , define
All weights are at most one. They can be negative outside when ; none of the following counts uses such negative incidences. Every owner incidence belongs to , since and .
Call an exception active if it belongs to at least two of the sets . An inactive exception belongs only to its owner's set. If is inactive with owner , then, for every ,
Every inactive exception is complete to its owner core, and the inactive exceptions with a common owner form a clique. To prove both assertions, consider a nonadjacent pair consisting either of such an and , or of two inactive exceptions with owner . In the first case has cross-degree below by the marking rule. Equation (1) therefore leaves at least common target nonneighbors in every other core, in both cases. Uniform representatives from these lists have expected target-edge count at most . They extend the pair to an independent -set, a contradiction.
For every the row-plus-exception cap is
where counts nonedges within . For , this follows from . Otherwise has order between and , and substitution in gives (2) exactly.
Let be the sum of weights over nonowner incidences , and let count these incidences. If is the active set, then
Put
Inactive exceptions have no owner holes, whence . Every nonedge within meets an active exception, because its inactive vertices form a clique. Thus
Summing (2), the owner weights sum to , so
Consequently
But
For , (3) is impossible, proving the span theorem. The argument pays for all owner holes and exceptional nonedges using the additional owner incidences, rather than bounding their product by .
4. Equality rigidity at the same quadratic threshold
Theorem. If , , , and all inherited caps hold, then partitions into cliques of order , with a matching between every pair.
For this was already proved. Otherwise (3) with forces . Every exception is inactive, so every core together with its assigned exceptions is a clique. These cliques partition vertices and each has order at most ; therefore each has order exactly . The cap on a full clique plus an outsider gives the asserted matching property.
Checks, improvement, and remaining gap
Both quadratic theorems received a separate audit of extraction, random transversals, negative-weight boundary cases, and the final incidence accounting. No additional external theorem is used beyond the self-contained finite partition lemma above.
These results supersede the previous span and equality thresholds. The eighth-power proof and the independent comparison remain as historical records of weaker methods and differing hypotheses.
A hypothetical overloaded core must now have
This still leaves an unbounded range in both main core reductions. Neither an unconditional reduction to this small-independence range nor a universal proof or counterexample has been obtained.