n67-e33da00a2a0818664a11
rows verifiedOne quadruple with a skew block at each of n=65, 67, 69 and 71, found on 2026-11-19 by the sampling search in research/starter/skew: the skew block is fixed in advance, invariant under a multiplier subgroup of odd order, and the other three blocks are sampled inside per-lag autocorrelation windows and joined exactly. The skew-Hadamard orders 260, 268, 276 and 284 are not among the open orders of baselines/skew-gaps.json; these are the board's first quadruples with a skew block at those n, without a claim of first existence. Quadruples with a skew block over the cyclic group were already published at all four orders: constructions of 1992 at 65 and 67 are recounted in [7], 69 is in [12], and 71 belongs to the family of [28] for prime powers q = 4n - 1 (q = 283). The search is not a census and says nothing about how many classes exist.
Original source attribution: vxvware
Submitted / imported by: vprusso · 2026-11-19
Board cell: With a skew block, Any
Entry file: quadruples/n67-e33da00a2a0818664a11.json · raw JSON
Rows
The four rows A, B, C, D of the canonical representative, with + for 1 and - for −1. The letter after each name is its symmetry: k skew, s symmetric, · neither. A skew row is listed first, so the Goethals–Seidel array of the rows in this order is a skew-Hadamard matrix of order 268. The verifier result below lists the same rows as canonical_rows in sorted order, in which the skew row is not first.
A k +-++-+-++--+++----+-+------++-+++-+---+--++++++-+-++++---++--+-+--+ B · +-------+-+-------+-+--++--+---++----+++++--+-----+-+--+++-+++-+-++ C · +---++++-+-++++-++-+++++----+-+-++-++--+-+-+---++-++-++--+---+++--+ D · +-+++---+-++-+-+--------+++-++-+--+---+--+----++---+++-+--++++-+--+
Picture
Notes
Found by a sampling search with the first block fixed in advance: a skew sequence invariant under the multiplier subgroup of order 33. The skew-Hadamard order 268 is not among the open orders of baselines/skew-gaps.json; this is the board's first quadruple with a skew block at n=67.
Checks and result
{
"id": "n67-e33da00a2a0818664a11",
"type": "quadruple",
"n": 67,
"group": {
"kind": "cyclic",
"invariant_factors": [
67
]
},
"tracks": [
"skew-sds",
"general"
],
"tier": "reproduced",
"checks": [
"integer-domain",
"exact-PAF",
"symmetry",
"canonical-form"
],
"equivalence": "equiv-v1",
"fingerprint": "gs-e33da00a2a0818664a113c621d5f8c9a1f31ed1325c9e3f369f3eebcff22d34a",
"canonical_rows": [
[
1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
1,
-1,
1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
1,
-1,
1,
-1,
-1,
1,
1,
-1,
-1,
1,
-1,
-1,
-1,
1,
1,
-1,
-1,
-1,
-1,
1,
1,
1,
1,
1,
-1,
-1,
1,
-1,
-1,
-1,
-1,
-1,
1,
-1,
1,
-1,
-1,
1,
1,
1,
-1,
1,
1,
1,
-1,
1,
-1,
1,
1
],
[
1,
-1,
-1,
-1,
1,
1,
1,
1,
-1,
1,
-1,
1,
1,
1,
1,
-1,
1,
1,
-1,
1,
1,
1,
1,
1,
-1,
-1,
-1,
-1,
1,
-1,
1,
-1,
1,
1,
-1,
1,
1,
-1,
-1,
1,
-1,
1,
-1,
1,
-1,
-1,
-1,
1,
1,
-1,
1,
1,
-1,
1,
1,
-1,
-1,
1,
-1,
-1,
-1,
1,
1,
1,
-1,
-1,
1
],
[
1,
-1,
1,
1,
-1,
1,
-1,
1,
1,
-1,
-1,
1,
1,
1,
-1,
-1,
-1,
-1,
1,
-1,
1,
-1,
-1,
-1,
-1,
-1,
-1,
1,
1,
-1,
1,
1,
1,
-1,
1,
-1,
-1,
-1,
1,
-1,
-1,
1,
1,
1,
1,
1,
1,
-1,
1,
-1,
1,
1,
1,
1,
-1,
-1,
-1,
1,
1,
-1,
-1,
1,
-1,
1,
-1,
-1,
1
],
[
1,
-1,
1,
1,
1,
-1,
-1,
-1,
1,
-1,
1,
1,
-1,
1,
-1,
1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
1,
1,
1,
-1,
1,
1,
-1,
1,
-1,
-1,
1,
-1,
-1,
-1,
1,
-1,
-1,
1,
-1,
-1,
-1,
-1,
1,
1,
-1,
-1,
-1,
1,
1,
1,
-1,
1,
-1,
-1,
1,
1,
1,
1,
-1,
1,
-1,
-1,
1
]
],
"row_sums": [
-13,
7,
1,
-7
],
"pattern": "··k·",
"flags": [],
"contributions": [
{
"id": "skew-pool-join-65-71",
"title": "Skew quadruples at n=65, 67, 69 and 71 by pool join",
"relation": "participant-result",
"publication_status": "unpublished",
"contributors": [
{
"name": "vxvware",
"github": "vxvware"
}
],
"summary": "One quadruple with a skew block at each of n=65, 67, 69 and 71, found on 2026-11-19 by the sampling search in research/starter/skew: the skew block is fixed in advance, invariant under a multiplier subgroup of odd order, and the other three blocks are sampled inside per-lag autocorrelation windows and joined exactly. The skew-Hadamard orders 260, 268, 276 and 284 are not among the open orders of baselines/skew-gaps.json; these are the board's first quadruples with a skew block at those n, without a claim of first existence. Quadruples with a skew block over the cyclic group were already published at all four orders: constructions of 1992 at 65 and 67 are recounted in [@djokovic1994five], 69 is in [@djokovic2023skew], and 71 belongs to the family of [@xia2005infinite] for prime powers q = 4n - 1 (q = 283). The search is not a census and says nothing about how many classes exist.",
"entry_ids": [
"n65-18dd1068d61602ed2938",
"n67-e33da00a2a0818664a11",
"n69-b5c03eb433648398afed",
"n71-dcfc88515e24802641f9"
],
"featured": false,
"stratum": "skew-sds",
"label": "Challenge result",
"position": 17,
"recorded_at": "2026-10-15"
}
]
}Sources
- vxvware pool join with a fixed skew block (research/starter/skew): run at n=67, 2026-11-19 · evidence/participants/pool-join/skew-67.txt
Submission JSON
{
"schema_version": 1,
"n": 67,
"group": {
"kind": "cyclic",
"invariant_factors": [
67
]
},
"submitted_by": "vprusso",
"submitted_at": "2026-11-19",
"discovered_by": [
"vxvware"
],
"sources": [
{
"title": "vxvware pool join with a fixed skew block (research/starter/skew): run at n=67, 2026-11-19",
"url": "https://github.com/vxvware/goethals-seidel-challenge/tree/fe66066ab3a1f9581841ca8cd972d6a5365c1217/research/starter/skew",
"commit": "fe66066ab3a1f9581841ca8cd972d6a5365c1217",
"source_path": "research/starter/skew",
"artifact": {
"path": "evidence/participants/pool-join/skew-67.txt",
"sha256": "4a4ed9177c93768e24c578644f284550fe61404419fe7e7f80d38a8213836c7d"
}
}
],
"equivalence": "equiv-v1",
"id": "n67-e33da00a2a0818664a11",
"type": "quadruple",
"rows": [
[
1,
1,
-1,
-1,
1,
-1,
1,
-1,
-1,
1,
1,
-1,
-1,
-1,
1,
1,
1,
1,
-1,
1,
-1,
1,
1,
1,
1,
1,
1,
-1,
-1,
1,
-1,
-1,
-1,
1,
-1,
1,
1,
1,
-1,
1,
1,
-1,
-1,
-1,
-1,
-1,
-1,
1,
-1,
1,
-1,
-1,
-1,
-1,
1,
1,
1,
-1,
-1,
1,
1,
-1,
1,
-1,
1,
1,
-1
],
[
-1,
-1,
1,
-1,
1,
-1,
1,
-1,
-1,
-1,
1,
1,
1,
1,
-1,
1,
1,
1,
1,
1,
-1,
1,
1,
1,
1,
1,
-1,
1,
1,
-1,
-1,
-1,
-1,
1,
-1,
-1,
-1,
1,
1,
1,
1,
1,
1,
-1,
-1,
-1,
-1,
1,
-1,
1,
1,
-1,
1,
-1,
-1,
1,
-1,
1,
1,
1,
-1,
-1,
1,
-1,
-1,
1,
1
],
[
1,
-1,
1,
-1,
1,
-1,
1,
1,
1,
-1,
-1,
1,
1,
1,
-1,
-1,
-1,
1,
1,
1,
-1,
-1,
1,
1,
-1,
1,
1,
1,
1,
-1,
1,
-1,
1,
1,
-1,
1,
-1,
1,
1,
-1,
1,
1,
-1,
-1,
1,
1,
1,
1,
1,
1,
-1,
-1,
-1,
-1,
-1,
1,
-1,
-1,
1,
-1,
-1,
1,
1,
-1,
-1,
1,
-1
],
[
-1,
1,
-1,
-1,
1,
1,
1,
-1,
-1,
-1,
1,
1,
-1,
-1,
1,
-1,
-1,
-1,
1,
1,
-1,
-1,
1,
-1,
1,
1,
1,
1,
1,
-1,
1,
-1,
-1,
1,
1,
1,
1,
1,
1,
-1,
1,
-1,
1,
1,
1,
1,
1,
1,
-1,
1,
-1,
1,
-1,
-1,
-1,
1,
-1,
-1,
1,
1,
1,
1,
1,
1,
-1,
1,
1
]
],
"notes": "Found by a sampling search with the first block fixed in advance: a skew sequence invariant under the multiplier subgroup of order 33. The skew-Hadamard order 268 is not among the open orders of baselines/skew-gaps.json; this is the board's first quadruple with a skew block at n=67."
}