T-192: Continuous-pose lower bound s(28)>1149/200 reported, by two of the source's programs

V0 C0 Lower bound recorded

2026-10-15 published · Hosono after Guzhou0806, Tokoharu, Levy · n=28

A rectangle measure of wand125/square-packing, published on 15 October 2026 and reported on no issue, is reported to prove s(28)>1149/200 = 5.745.

The certificate, continuous_n28_L5745, is a density of 884 uniform rectangles in D4 orbits, and no point mass, of total mass 2799999/100000 on K = [0, 573351/100000]^2. Every closed square of side B = 499/500 in K, at every centre and every angle, is reported to capture mass at least 1; 28 - 2799999/100000 = 1/100000 > 0, so no 28 squares of side B with disjoint interiors fit in K, and scaling by 1/B gives the claim, since L/B = 1149/200 exactly. The angle is covered continuously in t = tan(theta/2) up to 83/200, past tan(pi/8), by outward-rounded interval branch and bound, and the near-axis band by an exact check of a smaller axis-parallel square: T-145's method, on a new density at a new count.

The source reports two programs deciding it: sqverify_continuous v4, its continuous-angle mode of this repository's sqverify_fast crate, at all 86 root ranges (1,265,052,068 nodes, and a second complete run before publication), and contpose_check3 v1.2, a driver around Guzhou0806's T-145 kernels with their caps raised, at all 43 slabs (1,912,948,689 nodes). Each refused a control with every mass scaled by 985/1000.

The value is above T-142's reported 2297/400 = 5.7425 by 1/400, above the 287/50 that jlevy/squares#446 reported on 8 October, never adopted, by 1/200, and above T-107's 1147/200 = 5.735, the verified and reported lower bound at n=28, by 1/100.

Here the exact premises hold: the mass, the scalar chain, each program's near-axis side and margin, both logs against their summaries, the control's scaling, and both of check2's kernel inputs, regenerated from the candidate with the T-145 audit's builders, equal to the digests the source names. A diagnostic sample agreed with the source's logs: three of check1's roots, its control, two slabs on the reviewed T-145 kernel and the axis patch. A sample is not a replay. No complete replay has run here, so the value stands as reported.

Hosono (wand125) after Guzhou0806, Tokoharu and Levy, square-packing. The source's problems/square-lower-bounds/README.md states: "Parts of this work were produced with AI assistance under human direction." The commit names Claude as co-author, and the second program's specification is written for Codex.

Significance, composition and next rung
Significance
Scored as a claim: if a complete replay passes, it is the strongest lower bound on record at n=28, 1/400 above T-142 and 1/100 above T-107, the gap to the best known packing falling to about 0.0794. Its technique is T-145's continuous-pose cover, carried for the first time to a count other than 40 and decided at the source by two programs; a single-count raise on an existing technique and not a bound family, S3 as T-145, T-157 and T-142 are. V0/C0 says how far it stands, and the score gates nothing.
Composition
One reported certificate on its reported entry. The continuous cover is reported decided by two of the source's programs: sqverify_continuous, which keeps this repository's sqverify_fast interval, exact, axis, oracle and rotated modules byte for byte, and contpose_check3, which runs Guzhou0806's T-145 kernels with only their caps and receipt echo changed. The source says they share no code; they share none with each other, but each shares code this record already relies on, and both follow T-145's method. A replay here of the second on the reviewed T-145 kernel would be re-implemented sharing the producer's components; no independent method stands beside them.
Next rung
C3 needs a complete replay with controls whose refusals count and a review that reads this certificate. The nearer route runs the reviewed T-145 pose kernel on the input regenerated here over all 43 slabs, with the near-axis patch decided by sqverify-fast's exact sweep, priced at about 12.7 CPU-hours after a preparation slice; the other runs the source's sqverify_continuous over its 86 roots, about 13.3 CPU-hours, after the soundness review of its continuous mode (think-xzob, jlevy/squares#533). Both wait for an owner budget, in think-8eo1 and think-0arj under think-zb5w.
Unfinished confirmations
C3 (think-8eo1): a complete replay of the 43 slabs on the reviewed T-145 pose kernel from the input regenerated here, the near-axis patch decided by sqverify-fast at side 249499501/250000000, controls whose refusals count, and a review that reads this certificate. About 12.73 CPU-hours, up to about 30.43 (priced 2026-10-15).

C3 (think-0arj): a complete run of the source's sqverify_continuous v4 over all 86 roots with its controls, after the soundness review of its continuous mode. About 13.3 CPU-hours, up to about 16 (priced 2026-10-15).
Novelty
previously-published Present in an identified source

The case

Case record

n=28

5.735.824
567
5.2926.292
nn+1

Proven

5.73500≤s(28)≤5.824445

  • exact
  • rigid

Citation record n-028

lowerHosono after Tokoharu, Levy et al. 2026, GitHub (confirmed T-107)

upperEllsworth 2025, Squares in Squares (confirmed T-101, T-155)

Open

  • optimality

The case record

LowerUpper
Gap0.08944461…

Results on the case

15 results in the register on n=28, oldest first, each with what it established and how it stands now.

  1. 2005 published T-007

    s(n)≥min(⌈n⌉,n−2⌊n⌋+1+1) for 4≤n≤324

    V0 C1 Lower bound incomplete on this case, superseded by T-107

    Nagamochi · Nagamochi 2005 · source · register

  2. 2026-09-04 published T-085

    Nagamochi 2005, Lemma 1 is false for every container with a>3 and b>2

    V3 C3 Correction confirmed

    Karakuş; Vlasenko · Karakuş 2026 · chelokot Nagamochi counterexample 2026 · packet · register

  3. 2026-09-22 published T-047

    s(11)≥381/100; s(n)≥1377/250 for n=26…28; s(n)≥571/100 for n=29…31

    V3 C3 Lower bound confirmed superseded by T-107

    Tokoharu after Levy, Hosono, Stromquist, Nagamochi, Burns, Massaccesi · Tokoharu density 2026 · packet · source · review · register

  4. 2026-09-27 published T-045

    Rectangle-density lower bounds replayed at 15 counts in n=18…78

    V3 C3 Lower bound confirmed superseded by T-107

    Hosono after Tokoharu, Levy, Stromquist, Nagamochi, Burns, Massaccesi · wand125 rectangle bounds 2026 · packet · packet · source · review · register

  5. 2026-09-27 published T-046

    Rectangle-density lower bounds reported for 48 counts in n=18…95

    V0 C0 Lower bound recorded superseded by T-107

    Hosono after Tokoharu, Levy, Stromquist, Nagamochi, Burns, Massaccesi · wand125 rectangle bounds 2026 · wand125 rectangle bounds 2026-09-28 · packet · packet · register

  6. 2026-09-29 published T-058

    Rectangle-certificate ceiling α·UB(n) proved for n=1..100; B·UB(n) on 64 grid rows

    V3 C3 Method limit confirmed

    Hosono after Tokoharu, Daniel · wand125 tools 2026 · packet · register

  7. 2026-09-29 published T-083

    s(n)≥1/2+n−⌊n⌋+1/4 for every nonsquare 8≤n≤324

    V3 C3 Lower bound confirmed on this case, superseded by T-107

    Karakuş · Karakuş 2026 · source · register

  8. 2026-10-01 published T-068

    Rectangle-density lower bounds verified at 34 counts in n=19…95

    V3 C3 Lower bound confirmed on this case, superseded by T-107

    Hosono after Tokoharu, Levy, Stromquist, Nagamochi, Burns, Massaccesi · wand125 rectangle bounds 2026-10-01 · packet · source · review · register

  9. 2026-10-01 published T-074

    Rectangle-density lower bounds replayed at 31 counts in n=19…95

    V3 C3 Lower bound confirmed on this case, superseded by T-107

    Hosono after Tokoharu, Levy, Stromquist, Nagamochi, Burns, Massaccesi · wand125 rectangle bounds 2026-10-01 · packet · packet · source · review · register

  10. 2026-10-05 published T-101

    Exact certificates of 77 catalogue packings: s(n) ≤ S', 1.2e-16 to 9.8e-15 above each side

    V3 C3 Upper bound confirmed

    Daniel after Couzo, de Winter, Ellsworth, Levy · evand exact optima 2026-10-05 · packet · register

  11. 2026-10-06 published T-107

    Mixed rectangle-measure lower bound verified at n=28, on a declared net

    V3 C3 Lower bound confirmed

    Hosono after Tokoharu, Levy, Stromquist, Nagamochi, Burns, Massaccesi · wand125 mixed bounds check2 2026-10-06 · packet · source · review · register

  12. 2026-10-07 published T-124

    Reported non-strict local minima for 178 source configurations

    V0 C0 Restricted optimality recorded

    Daniel after Couzo · Daniel exact and local reports 2026 · packet · register

  13. 2026-10-10 published T-142

    Mixed rectangle-measure lower bound s(28)≥2297/400 reported, on a 2,073-direction declared net

    V0 C0 Lower bound recorded

    Hosono after Tokoharu and Levy · wand125 fine-net n28 n30 2026-10-10 · packet · register

  14. 2026-10-12 established T-155

    Exact regularized sides at 59 counts, 5.4e-20 to 9.9e-15 below the ceilings they replace

    V3 C3 Upper bound confirmed

    Levy after Daniel · register

  15. 2026-10-15 published T-192 this result

    Continuous-pose lower bound s(28)>1149/200 reported, by two of the source's programs

    V0 C0 Lower bound recorded

    Hosono after Guzhou0806, Tokoharu, Levy · wand125 continuous n28 2026-10-15 · packet · register