Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
On Small Folkman Graphs Arrowing K2 or K3
observation_11: A graph G with G → (3,3)^v gains G' → (3,3)^e when a vertex joined to every vertex of G is added; with Observation 12 and a 63-vertex C4-free polarity graph this gives F_e(3,3;W_5) at most 64, so that number exists.
reported_interval: Reports the previously known interval 21 <= F_e(3,3;4) <= 786 without claiming a new endpoint or an exact determination.
table_1: The paper's new values and bounds for vertex Folkman numbers with up to three colors and edge Folkman numbers F_e(3,3;H) avoiding J_4, ..., K_6, each from a computer search, with Corollary 10, F_e(3,3,3;8) at most 562.
table_2: The paper's new results F_v(2,3;C_4) = 17, 30 <= F_v(3,3;C_4) <= 63, F_e(3,3;W_5) <= 64, F_v(2,3,3,3;K_5) <= 32 and F_v(3,3,3,3;K_6) <= 30, each from a computer search or an explicit witness graph.
theorem_3: If the edge Folkman number avoiding the five-vertex complement of P2 ∪ P3 exists, a minimal arrowing graph for it is K4-free, has every edge in at least two triangles and minimum degree at least 8; its Corollary 4 bounds that number below by F_e(3,3;4).
Zohair Raza Hassan, Stanisław Radziszowski, and Steven Van Overberghe, On Small Folkman Graphs Arrowing or , arXiv:2605.16542v1 (15 May 2026). The supplied record establishes this preprint version; it does not establish acceptance or publication. The arXiv record names arXiv's non-exclusive distribution license (arXiv:2605.16542), every other right reserved.
Edition read. The copy read for this card is the arXiv v1 PDF, 21 physical pages. The title, authors, and version are on physical p. 1. The prose on physical and printed p. 3 reports the interval $21\leq F_e(3,3;4)\leq786$; Table 1 repeats it on physical and printed p. 4.
Digest. The paper gives new values and bounds for vertex and edge Folkman numbers arrowing and with up to four colors while avoiding and for , and , collected in Table 1 (p. 4) and Table 2 (p. 5). Nearly all are computational: exhaustive generation with subgraph filters, extension of smaller graphs, a semi-polycirculant graph generator, generation of locally linear graphs, and modification of known special graphs (Sections 2--4, pp. 5--17). Through Lemma 9, the bound gives Corollary 10, and (p. 13). Observation 11 (p. 16) turns a -arrowing graph into a -arrowing one by adding a universal vertex, and with Observation 12 gives (p. 17), answering an existence question of Hassan, Jiang, Narváez, Radziszowski and Xu. Theorem 3 (pp. 3--4, proved on pp. 17--18) lists properties of a minimal witness for , if one exists, among them that it is -free; its Corollary 4 is , and the paper closes by asking whether that number exists (Problem 1, p. 18).
For Problem 582 the paper proves no new bound on . It reports the known interval
on p. 3 and in Table 1 on p. 4, attributing the endpoints jointly to Bikov--Nenov and Lange--Radziszowski--Xu; it does not determine the exact minimum and does not reproduce either endpoint proof or the MAX-CUT/SDP certificate (reported interval).
Source: https://arxiv.org/abs/2605.16542.
Bears on. #582: the reported interval restates prior bounds on the least order of a graph of the kind the problem asks for, and Theorem 3 (item 1) shows that a minimal witness for , if one exists, is such a graph, whence its Corollary 4, . Neither adds to the existence that Folkman's theorem settled.
Results.
- Theorem 3 with Corollary 4 and Problem 1: pp. 3--4, 17--18.
- Table 1 with Corollary 10: pp. 4 and 13.
- Table 2: p. 5.
- Observation 11 with Observation 12 and : pp. 16--17.
- Reported interval: pp. 3--4 retain as previously known context.
Read status. Claims checked for the statements on these pages, read on the page images of the arXiv v1 PDF; the computer searches and witness graphs were not rerun or checked. Nothing is independently reviewed.
No file of this source is held: no license on record permits its redistribution, and the card cites the edition it names above.