/teal-sea
teal-sea / zeta-labstate of record · compiled 28 Sep 2026 · revision e4945c4 · source

Library · hunts/r_322dae/RESULTS.md

Hunt R-322DAE Results: Krenn-Gu 8x3 Orbit Census and Next Exact Frontier

1,826 words · 130 lines · source

Executive Summary

This hunt settles the orbit-census layer for the Krenn-Gu 8x3 instance ($n=8$ vertices, $d=3$ colors).

  1. 6x3 Baseline Checksum Verified:
  2. 15 perfect matchings ($(6-1)!! = 15$).
  3. $15^3 = 3375$ monochromatic matching triples.
  4. Exact group action under $S_6 \times S_3$ ($|G| = 4320$).
  5. Burnside orbit count: exactly 8.
  6. 8 disjoint orbits with sizes $[15, 90, 120, 270, 360, 360, 1080, 1080]$, summing to 3375.
  7. Fixing the first matching $M_0$ and quotienting by its stabilizer $H = S_2 \wr S_3$ ($|H| = 48$) yields 16 pair orbits (12 under $H \times S_2$).
  1. 8x3 Orbit Census Settled:
  2. 105 perfect matchings ($(8-1)!! = 105$).
  3. $105^3 = 1157625$ monochromatic matching triples.
  4. Exact group action under $S_8 \times S_3$ ($|G| = 40320 \times 6 = 241920$).
  5. Burnside orbit count: exactly 31.
  6. 31 disjoint orbits covering all 1157625 triples with zero overlap and zero unvisited triples ($\sum |Orbit_i| = 1157625$).
  7. For each orbit $i$, the orbit-stabilizer theorem $|Orbit_i| \times |Stab_i| = 241920$ holds exactly.
  8. Fixing the first matching $M_0$ and quotienting by its wreath-product stabilizer $H = S_2 \wr S_4$ ($|H| = 384$) partitions the 11025 pairs $(M_1, M_2)$ into 86 orbits (57 orbits under $H \times S_2$).

Total runtime for the census was 13.8 seconds with peak memory well below 50 MB, far within the 45 minute and 8 GB budget limits.


1. The 6x3 Baseline Checksum

Under $G = S_6 \times S_3$ ($|G| = 4320$):

Orbit IDCanonical TripleDistinct MatchingsOrbit SizeStabilizer Size2-Factor Cycle Structures
0(0, 0, 0)115288((2, 2, 2), (2, 2, 2), (2, 2, 2))
1(0, 0, 1)227016((2, 2, 2), (4, 2), (4, 2))
2(0, 0, 2)236012((2, 2, 2), (6,), (6,))
3(0, 1, 2)39048((4, 2), (4, 2), (4, 2))
4(0, 1, 3)310804((4, 2), (4, 2), (6,))
5(0, 1, 4)310804((4, 2), (6,), (6,))
6(0, 2, 7)336012((6,), (6,), (6,))
7(0, 2, 8)312036((6,), (6,), (6,))

2. The 8x3 Orbit Census Table

Under $G = S_8 \times S_3$ ($|G| = 241920$):

OrbitCanonical IndicesDistOrbit SizeStabMult (3,2,1)Components2-Factor Cycles $((M_0,M_1), (M_0,M_2), (M_1,M_2))$
0(0, 0, 0)11052304(4, 0, 0)(2, 2, 2, 2)((2, 2, 2, 2), (2, 2, 2, 2), (2, 2, 2, 2))
1(0, 0, 1)2378064(2, 2, 2)(4, 2, 2)((2, 2, 2, 2), (4, 2, 2), (4, 2, 2))
2(0, 0, 4)21008024(1, 3, 3)(6, 2)((2, 2, 2, 2), (6, 2), (6, 2))
3(0, 0, 16)2378064(0, 4, 4)(4, 4)((2, 2, 2, 2), (4, 4), (4, 4))
4(0, 0, 19)21512016(0, 4, 4)(8,)((2, 2, 2, 2), (8,), (8,))
5(0, 1, 2)31260192(2, 0, 6)(4, 2, 2)((4, 2, 2), (4, 2, 2), (4, 2, 2))
6(0, 1, 3)3302408(1, 2, 5)(6, 2)((4, 2, 2), (4, 2, 2), (6, 2))
7(0, 1, 5)3302408(1, 1, 7)(6, 2)((4, 2, 2), (6, 2), (6, 2))
8(0, 1, 15)3756032(0, 4, 4)(4, 4)((4, 2, 2), (4, 2, 2), (4, 4))
9(0, 1, 17)3756032(0, 2, 8)(4, 4)((4, 2, 2), (4, 4), (4, 4))
10(0, 1, 18)31209602(0, 3, 6)(8,)((4, 2, 2), (6, 2), (8,))
11(0, 1, 20)3604804(0, 2, 8)(8,)((4, 2, 2), (8,), (8,))
12(0, 1, 52)3604804(0, 2, 8)(8,)((4, 2, 2), (4, 4), (8,))
13(0, 1, 58)3302408(0, 2, 8)(8,)((4, 2, 2), (8,), (8,))
14(0, 4, 8)31008024(1, 0, 9)(6, 2)((6, 2), (6, 2), (6, 2))
15(0, 4, 13)3336072(1, 0, 9)(6, 2)((6, 2), (6, 2), (6, 2))
16(0, 4, 16)3604804(0, 2, 8)(8,)((4, 4), (6, 2), (6, 2))
17(0, 4, 17)31209602(0, 1, 10)(8,)((4, 4), (6, 2), (8,))
18(0, 4, 21)31209602(0, 2, 8)(8,)((6, 2), (6, 2), (8,))
19(0, 4, 23)31209602(0, 1, 10)(8,)((6, 2), (8,), (8,))
20(0, 4, 27)3604804(0, 1, 10)(8,)((6, 2), (8,), (8,))
21(0, 4, 28)3604804(0, 1, 10)(8,)((6, 2), (8,), (8,))
22(0, 4, 29)32016012(0, 3, 6)(8,)((6, 2), (6, 2), (6, 2))
23(0, 16, 32)31260192(0, 0, 12)(4, 4)((4, 4), (4, 4), (4, 4))
24(0, 16, 35)3302408(0, 0, 12)(8,)((4, 4), (8,), (8,))
25(0, 16, 52)3504048(0, 0, 12)(8,)((4, 4), (4, 4), (4, 4))
26(0, 16, 53)31512016(0, 0, 12)(8,)((4, 4), (4, 4), (8,))
27(0, 16, 55)31512016(0, 0, 12)(8,)((4, 4), (8,), (8,))
28(0, 16, 56)3302408(0, 0, 12)(8,)((4, 4), (8,), (8,))
29(0, 19, 38)3604804(0, 0, 12)(8,)((8,), (8,), (8,))
30(0, 19, 43)3403206(0, 0, 12)(8,)((8,), (8,), (8,))

3. Structural Decomposition and Stabilizer Quotients

Fixing the first matching $M_0$ (without loss of generality, $M_0 = \{(0,1), (2,3), (4,5), (6,7)\}$) restricts the vertex permutation group to the wreath product: $$H = Stab_{S_8}(M_0) = S_2 \wr S_4 = B_4 \quad (|H| = 2^4 \times 4! = 384)$$

The remaining search space consists of $105^2 = 11025$ matching pairs $(M_1, M_2)$.

These 31 classes give an exhaustive outer branch cover after choosing one supported monochromatic matching in each colour. They do not quotient the 252 edge-weight variables or the 6,561 polynomial equations: a branch still carries the full equation system. Hunt R-044DD2 records this scope correction and uses the classes only as target-support skeletons.


4. What Was Chosen and Why


5. What Could Not Be Settled


Loose threads

  1. Exact support and algebraic sieving over the 31 outer branches:
  2. What it is: Require the twelve diagonal entries of one target-matching representative, then impose the full 6,561-equation support necessities and learn exact algebraic no-goods from returned supports.
  3. Why it matters: This preserves the valid symmetry cover without assuming that a hypothetical witness is itself symmetric.
  4. First step: Run the support frontier and signed-Laurent sieve in Hunt R-044DD2.