MinMax Arena
中/EN
BREAKTHROUGH REVIEW

Breakthrough review.

A numerical lead is evidence to review, not a conclusion. Pending entries pass the arena verifier; certified entries additionally rule out source, normalisation and published-precision artefacts.

Verified breakthroughs39
Pending candidates47
Problem families9
CLAIM POLICY

One certification standard

  1. 01A traceable external benchmarkThere must be a public source and an explicit comparison value. A baseline found by this site is not an external record.
  2. 02A complete construction verifiesThe complete answer is rescored by the submission verifier. The verified entry is pinned to that exact score and contributor; it does not transfer with the current record.
  3. 03Rule out precision artefactsTruncated and rounded values are expanded to their most conservative interval, together with source-coordinate precision, solver tolerance and reconstruction error. A last-digit floating-point tweak without a rigorous error bound remains only a candidate; verification requires either a proof or a margin beyond every identified uncertainty.

A verified breakthrough is not a proof of global optimality. It means a uniformly reviewed construction genuinely improves on a traceable external best-known construction. A player can earn the 3-point award only once per sub-problem, and keeps it after the record changes hands.

UNDER REVIEW

Numerical leads under review

These complete constructions pass the arena verifier, but their original sources, unpublished digits or numerical error bounds still require review. They receive no breakthrough points here.

Current verifiable construction for Lighting a unit square, n = 28
01 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 28

Published best known252.447+
→
Arena candidate268.159996
Numerical lead+6.223854%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 252.447+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 252.429969, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 28
02 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 28

Published best known252.447+
→
Arena candidate268.155641
Numerical lead+6.222129%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 252.447+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 252.429969, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 20
03 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 20

Published best known165.272+
→
Arena candidate174.540507
Numerical lead+5.607393%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 165.272+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 165.261482, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 20
04 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 20

Published best known165.272+
→
Arena candidate174.538563
Numerical lead+5.606217%
Candidate contributor
今天#84
Solution method
Human
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 165.272+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 165.261482, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 27
05 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 27

Published best known242.055+
→
Arena candidate252.071978
Numerical lead+4.137876%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 242.055+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 242.038739, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 27
06 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 27

Published best known242.055+
→
Arena candidate252.066633
Numerical lead+4.135668%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 242.055+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 242.038739, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 18
07 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 18

Published best known141.556+
→
Arena candidate147.340099
Numerical lead+4.08535%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 141.556+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 141.549168, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 18
08 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 18

Published best known141.556+
→
Arena candidate147.340097
Numerical lead+4.085349%
Candidate contributor
今天#84
Solution method
Human
Submitted
6d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 141.556+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 141.549168, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 19
09 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 19

Published best known152.854+
→
Arena candidate158.869529
Numerical lead+3.934794%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 152.854+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 152.836737, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 19
10 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 19

Published best known152.854+
→
Arena candidate158.868906
Numerical lead+3.934386%
Candidate contributor
今天#84
Solution method
Human
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 152.854+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 152.836737, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 19
11 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 19

Published best known152.854+
→
Arena candidate158.868738
Numerical lead+3.934276%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 152.854+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 152.836737, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 30
12 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 30

Published best known285.345+
→
Arena candidate294.523343
Numerical lead+3.216216%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Shaoheng Lai and Codex (August 2026) on Erich Friedman's page as 285.345+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 285.33045, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 30
13 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 30

Published best known285.345+
→
Arena candidate294.519372
Numerical lead+3.214824%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Shaoheng Lai and Codex (August 2026) on Erich Friedman's page as 285.345+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 285.33045, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 35
14 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 35

Published best known347.195+
→
Arena candidate357.334366
Numerical lead+2.92007%
Candidate contributor
今天#84
Solution method
Human
Submitted
6d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Shaoheng Lai and Codex (August 2026) on Erich Friedman's page as 347.195+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 347.191734, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 26
15 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 26

Published best known234.759+
→
Arena candidate239.68973
Numerical lead+2.099902%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 234.759+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 234.742765, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 26
16 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 26

Published best known234.759+
→
Arena candidate239.688299
Numerical lead+2.099292%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 234.759+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 234.742765, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 34
17 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 34

Published best known337.906+
→
Arena candidate344.245377
Numerical lead+1.875776%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Shaoheng Lai and Codex (August 2026) on Erich Friedman's page as 337.906+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 337.902847, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 34
18 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 34

Published best known337.906+
→
Arena candidate344.242054
Numerical lead+1.874792%
Candidate contributor
今天#84
Solution method
Human
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Shaoheng Lai and Codex (August 2026) on Erich Friedman's page as 337.906+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 337.902847, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 10
19 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 10

Published best known62.408+
→
Arena candidate63.548774
Numerical lead+1.826297%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 62.408+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 62.403068, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 10
20 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 10

Published best known62.408+
→
Arena candidate63.548773
Numerical lead+1.826296%
Candidate contributor
今天#84
Solution method
Human
Submitted
6d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 62.408+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 62.403068, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 10
21 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 10

Published best known62.408+
→
Arena candidate63.548767
Numerical lead+1.826286%
Candidate contributor
215590053#88
Solution method
Human
Submitted
6d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 62.408+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 62.403068, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 15
22 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 15

Published best known113.191+
→
Arena candidate114.942022
Numerical lead+1.546065%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 113.191+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 113.185559, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 15
23 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 15

Published best known113.191+
→
Arena candidate114.942016
Numerical lead+1.54606%
Candidate contributor
hadsajnc#146
Solution method
Human
Submitted
2d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 113.191+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 113.185559, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 15
24 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 15

Published best known113.191+
→
Arena candidate114.942009
Numerical lead+1.546054%
Candidate contributor
今天#84
Solution method
Human
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 113.191+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 113.185559, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 25
25 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 25

Published best known225.100+
→
Arena candidate228.072119
Numerical lead+1.319905%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 225.100+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 225.085358, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 25
26 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 25

Published best known225.100+
→
Arena candidate228.066424
Numerical lead+1.317375%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 225.100+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 225.085358, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 24
27 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 24

Published best known215.573+
→
Arena candidate218.381153
Numerical lead+1.302176%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 215.573+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 215.555155, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 24
28 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 24

Published best known215.573+
→
Arena candidate218.375937
Numerical lead+1.299756%
Candidate contributor
今天#84
Solution method
Human
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 215.573+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 215.555155, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 24
29 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 24

Published best known215.573+
→
Arena candidate218.375555
Numerical lead+1.299579%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 215.573+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 215.555155, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 17
30 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 17

Published best known133.907+
→
Arena candidate134.777372
Numerical lead+0.649231%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 133.907+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 133.899583, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 17
31 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 17

Published best known133.907+
→
Arena candidate134.777334
Numerical lead+0.649202%
Candidate contributor
hadsajnc#146
Solution method
Human
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 133.907+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 133.899583, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 17
32 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 17

Published best known133.907+
→
Arena candidate134.777333
Numerical lead+0.649202%
Candidate contributor
今天#84
Solution method
Human
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 133.907+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 133.899583, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 17
33 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 17

Published best known133.907+
→
Arena candidate134.777195
Numerical lead+0.649099%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Satoshi Kishimoto and GPT-5.6 Sol Pro (July 2026) on Erich Friedman's page as 133.907+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 133.899583, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 31
34 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 31

Published best known305.298+
→
Arena candidate305.872231
Numerical lead+0.187761%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
11h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Shaoheng Lai and Codex (August 2026) on Erich Friedman's page as 305.298+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 305.28871, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 31
35 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 31

Published best known305.298+
→
Arena candidate305.851239
Numerical lead+0.180885%
Candidate contributor
邱仲普#72
Solution method
AI · MoonshotAI: Kimi K3
Submitted
5d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Shaoheng Lai and Codex (August 2026) on Erich Friedman's page as 305.298+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 305.28871, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Lighting a unit square, n = 5
36 · P51PENDING · UNVERIFIED

Lighting a unit square · n = 5

Published best known22.213+
→
Arena candidate22.252209
Numerical lead+0.172004%
Candidate contributor
blebub#215
Solution method
AI · Anthropic: Claude Sonnet 5
Submitted
4d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

published by Erich Friedman (August 2026) on Erich Friedman's page as 22.213+, meaning at least that much. The page gives figures and no coordinates; the configuration this site reconstructed from the pictured topology and symmetries verifies at 22.212249, which is the site's reference answer and not the record. Optimality is open

Current verifiable construction for Packing circles of radius 1,2,…,n into a circle, n = 57
37 · P06PENDING · UNVERIFIED

Packing circles of radius 1,2,…,n into a circle · n = 57

Published best known[268.35813003045, 268.35813003055]
→
Arena candidate268.23567911
Numerical lead+0.04563%
Candidate contributor
邱仲普#72
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
7h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

Packomania's ccin ratio table and public coordinates (Eckard Specht); the source precision is preserved and the coordinates are conservatively repaired to the site's nine-place grid

Current verifiable construction for Tilted equal squares in a circle, n = 31
38 · P05PENDING · UNVERIFIED

Tilted equal squares in a circle · n = 31

Published best known[0.2834692095744560, 0.2834700131246617]
→
Arena candidate0.2835676
Numerical lead+0.034426%
Candidate contributor
游侠#116
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
9h ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

Haowei Lin found the r = 3.52771+ packing in July 2026 (radius of the smallest circle around unit squares). The site digitised the squares' positions and orientations from the public figure, solved that fixed separating-axis structure and repaired it inward to a nine-decimal certificate scoring 0.080355247477381876; the truncated source interval is compared separately from the certificate, and optimality is open

Current verifiable construction for The most separated family of subspaces, d = 5, n = 19
39 · P64PENDING · UNVERIFIED

The most separated family of subspaces · d = 5, n = 19

Published best known[1.1761278309, 1.1761278310]
→
Arena candidate1.176128991
Numerical lead+0.000099%
Candidate contributor
邱仲普#72
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
2d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

the public Grassmannian configuration archive maintained by Henry Cohn (converted from the source-precision squared chordal distance D; the site's nine-place certificate is listed separately)

Current verifiable construction for The most separated family of subspaces, d = 5, n = 20
40 · P64PENDING · UNVERIFIED

The most separated family of subspaces · d = 5, n = 20

Published best known[1.1619417514, 1.1619417515]
→
Arena candidate1.161941835
Numerical lead+0.000007%
Candidate contributor
邱仲普#72
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
2d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

the public Grassmannian configuration archive maintained by Henry Cohn (converted from the source-precision squared chordal distance D; the site's nine-place certificate is listed separately)

Current verifiable construction for The most separated family of subspaces, d = 6, n = 15
41 · P64PENDING · UNVERIFIED

The most separated family of subspaces · d = 6, n = 15

Published best known[1.4285668598, 1.4285668599]
→
Arena candidate1.428566861
Numerical lead+<0.000001%
Candidate contributor
邱仲普#72
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
2d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

the public Grassmannian configuration archive maintained by Henry Cohn (converted from the source-precision squared chordal distance D; the site's nine-place certificate is listed separately)

Current verifiable construction for Covering a regular pentagon with n equal discs, n = 10
42 · P75PENDING · UNVERIFIED

Covering a regular pentagon with n equal discs · n = 10

Site reconstruction of published figure0.174314361048804
→
Arena candidate0.174311702883821
Source value unresolved—
Candidate contributor
邱仲普#72
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
3d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

The public vector construction in Figure 9 of Yanchao Liu (2022), reconstructed point by point and verified here at 0.174314361048804. Table 1's 18.338 conflicts with the Figure 9 reconstruction, and the PDF drawing coordinates are rounded. The original radius is unresolved: this reconstruction cannot establish a match or advance over the literature and remains a review item. The paper's best lower bound still leaves a 10.24% gap, so optimality is open.

Current verifiable construction for Covering a regular pentagon with n equal discs, n = 9
43 · P75PENDING · UNVERIFIED

Covering a regular pentagon with n equal discs · n = 9

Site reconstruction of published figure0.181871774083792
→
Arena candidate0.181862304185616
Source value unresolved—
Candidate contributor
邱仲普#72
Solution method
AI · OpenAI: GPT-5.6 Sol
Submitted
3d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

The public vector construction in Figure 9 of Yanchao Liu (2022), reconstructed point by point and verified here at 0.181871774083792. Table 1's 19.377 conflicts with the Figure 9 reconstruction, and the PDF drawing coordinates are rounded. The original radius is unresolved: this reconstruction cannot establish a match or advance over the literature and remains a review item. The paper's best lower bound still leaves a 6.35% gap, so optimality is open.

Current verifiable construction for Covering a regular pentagon with n equal discs, n = 10
44 · P75PENDING · UNVERIFIED

Covering a regular pentagon with n equal discs · n = 10

Site reconstruction of published figure0.174314361048804
→
Arena candidate0.174311703583512
Source value unresolved—
Candidate contributor
hadsajnc#146
Solution method
Human
Submitted
7d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

The public vector construction in Figure 9 of Yanchao Liu (2022), reconstructed point by point and verified here at 0.174314361048804. Table 1's 18.338 conflicts with the Figure 9 reconstruction, and the PDF drawing coordinates are rounded. The original radius is unresolved: this reconstruction cannot establish a match or advance over the literature and remains a review item. The paper's best lower bound still leaves a 10.24% gap, so optimality is open.

Current verifiable construction for Covering a regular pentagon with n equal discs, n = 9
45 · P75PENDING · UNVERIFIED

Covering a regular pentagon with n equal discs · n = 9

Site reconstruction of published figure0.181871774083792
→
Arena candidate0.181862304660883
Source value unresolved—
Candidate contributor
hadsajnc#146
Solution method
Human
Submitted
7d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

The public vector construction in Figure 9 of Yanchao Liu (2022), reconstructed point by point and verified here at 0.181871774083792. Table 1's 19.377 conflicts with the Figure 9 reconstruction, and the PDF drawing coordinates are rounded. The original radius is unresolved: this reconstruction cannot establish a match or advance over the literature and remains a review item. The paper's best lower bound still leaves a 6.35% gap, so optimality is open.

Current verifiable construction for Covering a regular pentagon with n equal discs, n = 10
46 · P75PENDING · UNVERIFIED

Covering a regular pentagon with n equal discs · n = 10

Site reconstruction of published figure0.174314361048804
→
Arena candidate0.174311893039758
Source value unresolved—
Candidate contributor
NUE_13#31
Solution method
Human
Submitted
8d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

The public vector construction in Figure 9 of Yanchao Liu (2022), reconstructed point by point and verified here at 0.174314361048804. Table 1's 18.338 conflicts with the Figure 9 reconstruction, and the PDF drawing coordinates are rounded. The original radius is unresolved: this reconstruction cannot establish a match or advance over the literature and remains a review item. The paper's best lower bound still leaves a 10.24% gap, so optimality is open.

Current verifiable construction for Covering a regular pentagon with n equal discs, n = 9
47 · P75PENDING · UNVERIFIED

Covering a regular pentagon with n equal discs · n = 9

Site reconstruction of published figure0.181871774083792
→
Arena candidate0.181862735625360
Source value unresolved—
Candidate contributor
NUE_13#31
Solution method
Human
Submitted
8d ago

Not certified: For now we know only that the construction is exactly feasible here and numerically crosses the recorded comparison boundary. It is not claimed as a genuine literature breakthrough until the external evidence review is complete.

External source ↗Inspect candidate construction →

The public vector construction in Figure 9 of Yanchao Liu (2022), reconstructed point by point and verified here at 0.181871774083792. Table 1's 19.377 conflicts with the Figure 9 reconstruction, and the PDF drawing coordinates are rounded. The original radius is unresolved: this reconstruction cannot establish a match or advance over the literature and remains a review item. The paper's best lower bound still leaves a 6.35% gap, so optimality is open.