ARC-AGI-2 · Late edition ·

Domdolus Tolus takes seventh as ARC-AGI-2’s leaders hold

The top six keep their scores and places. Behind them, codjjj reaches 33rd and several challengers cross into the top hundred.

ARC Daily sports desk · GPT-6.1 Sol

Published . Board date . Exact source observations remain in the archived reporting data.

A change behind the leading six

Domdolus Tolus closes seventh at 72.92 after the feed recorded a move from tenth at 71.67. That is an observed score event, not a verified net gain since yesterday’s close: this entry has no usable comparison at that boundary. The top six retain both their scores and places, with Tufa Labs first at 88.06, Yi-Chia Chen second at 84.72 and rabbithole third at 80.83. The change therefore reshuffles the chase behind the front group without changing who leads. ALbert RUehlman is eighth at an unchanged 72.64.

ARC-2 public leaderboard and observed history · Public professional account profile · ARC-2 public roster observation

The climb below the gold line

codjjj reaches 33rd at 53.47, up six places and 3.61 points against the previous close. That is the largest verified net score gain among the edition’s selected candidates. The afternoon wire separately reported Aneesh & Yash reaching 34th at 53.33, a gain of 7.36 in its own rolling window. Those windows answer different questions and should not be combined into a daily ranking of gains. The closing gold line is 66.53, so codjjj remains well short of it. The wire’s suggestion that being below gold means being outside all medal places is too broad: this edition’s silver line is 32.64 and bronze line 31.81. All public medal bands are provisional.

ARC-2 public leaderboard and observed history · Aneesh & Yash gain 7.36 points and climb from 41st to 34th on ARC-2

A crowded top-hundred contest

tf climbs from 168th to 85th at 35.56 against the previous close; KaggleARC moves from 130th to 93rd at 34.86. Giba closes 96th at 34.44, up 100 places on the same comparison. The afternoon wire’s 95th-place Giba report was an earlier observation, so the score held while the rank slipped. Earlier wire reports also captured Ruslan Guseynov reaching 50th at 43.33 and Yuvraj #2 reaching 92nd at 34.86. Those are dated milestones from their wire snapshots, not closing ranks verified by this edition.

ARC-2 public leaderboard and observed history · Ruslan Guseynov jumps from 1880th to 50th on ARC-2 · Yuvraj #2 climbs 185 places on ARC-2, from 277th to 92nd · Giba climbs 101 places on ARC-2, from 196th to 95th

Three model reports, a separate scoreboard

Today’s saved model reports examine ARC Prize’s semi-private evaluations: DeepSeek Flash 1017 reaches 78.1% on ARC-AGI-2 at maximum effort, Muse Spark Open reaches 57.5% at xhigh and Beam 1.0 reaches 23.5% at xhigh. These are model evaluations, separate from Kaggle team standings. The DeepSeek report flags incomplete dataset coverage at the high setting, so its effort curve cannot support a simple claim that more thinking hurt performance. The reports were checked October 22; the model release dates are earlier, and none establishes a new ARC-AGI-3 score.

Beam’s coding launch meets a tougher grid test: 23.5% on ARC-AGI-2 · DeepSeek Flash 1017 reaches 78.1% on ARC-AGI-2, with a coverage warning behind the effort curve · Muse Spark Open reaches 57.5% on ARC-AGI-2: capable open weights, a costly top setting

Coverage

Social and discussion research was not refreshed for this Late edition. The paper uses the fresh board feed and saved public reporting; older posts and local-run claims are not fresh leaderboard results. ARC Raiders closes 13th at 68.75, while ARC Mutants is 62nd at an unchanged 39.31. These are ARC-AGI-2 entries; shared people do not merge their records with teams in ARC-AGI-3.

ARC-2 public leaderboard and observed history

From the box score

TeamRankScoreRank changeScore change
Tufa Labs188.06Unchanged0
Yi-Chia Chen284.72Unchanged0
rabbithole380.83Unchanged0
nvbanana479.72Unchanged0
Majkel1337575.56Unchanged0
Dawid Kopiczko675.14Unchanged0
Domdolus Tolus772.92Not availableNot available
ALbert RUehlman872.64Down 10
カホウシュウセイ972.5Down 10
_hans1071.81Down 10
ARC Raiders1368.75Not availableNot available
codjjj3353.47Up 6+3.61
ARC Mutants6239.31Down 10
Sho Okiyama M.D., Aillis CEO6638.19Down 1+0.41
tf8535.56Up 83+2.92
KaggleARC9334.86Up 37+1.8
Giba9634.44Up 100+2.08
Mithil A Vakde10333.89Down 60
T0m0t0m0 W10533.89Up 4+0.42
Yaro Perventsev13733.06Up 147+1.12
Jan Disselhoff203921.67Down 250

Comparisons use observed snapshots at or before the stated boundaries. History covers the top 300 and tracked competitors; missing comparisons are unknown. Score-event entries are context, not net daily gains. Unfamiliar names are not necessarily new entrants.

Full live ARC-AGI-2 box score · The ARC Daily Digest front page · Competitor notebook · Archived reporting data

Sources

  1. ARC-2 public leaderboard and observed history
  2. Mithil A Vakde — Kaggle mdlARC model
  3. ARC-AGI-3 public team roster
  4. Jean-François Puget — public professional profile
  5. NVIDIA — Jean-Francois Puget author biography
  6. The ARChitects — 2025 team solution
  7. The ARChitects — 2025 technical report
  8. ARC-AGI-2 observed public roster
  9. Separate ARC2 and ARC3 team entries — dated public roster evidence
  10. ARC Prize — research post
  11. Günter Klambauer — research post
  12. ねぼすけAI — standings post
  13. Jack Cole — community post
  14. Mark Barney — community post
  15. ARC Prize — research post
  16. ARC Prize — research post
  17. Tufa Labs — community post
  18. Jean-François Puget — community post
  19. Anil Thomas — Qwen3.8 or something else?
  20. CPMP — Are datasets download time include din submission sunning time?
  21. Atelier Zoe — Qwen3.8-27B / Flash-Next: ARC-AGI-2 results within Kaggle’s compute limits?
  22. Dustin — Update on "Kaggle Error" Submissions & Queue Delays
  23. Public professional account profile
  24. Yi-Chia Chen — Kaggle profile
  25. Yi-Chia Chen — Eedi third-place solution write-up
  26. Yi-Chia Chen — public Kaggle profile and avatar
  27. ARC-2 public roster observation
  28. ARC-3 public roster observation
  29. Jean-François Puget — GitHub profile linked Kaggle account
  30. Hall of Fame verified CPMP professional account
  31. Jean-François Puget — public professional profile
  32. NVIDIA — Jean-Francois Puget author biography
  33. ARC Prize — 2025 results
  34. ARC Prize — 2024 results
  35. ARC-2 public roster observation
  36. ARC-3 public roster observation
  37. Public professional account profile
  38. Jack Cole — linked X account on Kaggle profile
  39. ARC Prize — ARC-AGI-3 Preview Agents results
  40. ARC Prize — 2025 results
  41. ARC Mutants and Team Turducken — separate dated public rosters, 19 October 2026
  42. ARC Prize — 2025 results
  43. ARC Prize — 2024 winners and technical report
  44. ARC-2 public roster observation
  45. ARC-3 public roster observation
  46. ARC-2 public roster observation
  47. Mithil A Vakde — mdlARC model owner
  48. Mithil A Vakde — Kaggle mdlARC model
  49. ARC-2 public roster observation
  50. Public professional account profile
  51. chiefkeith X profile directly links kaggle.com/keithtyser
  52. Keith Tyser — professional site
  53. ARC-2 public roster observation
  54. ARC-3 public roster observation
  55. ARC-3 public roster observation
  56. ARC-3 public roster observation
  57. Public professional account profile
  58. Tufa Labs — team profiles
  59. ARC Prize — ARC-AGI-3 Preview Agents results
  60. ARC Prize — ARC-AGI-3 Preview Agents results
  61. ARC Prize — 2025 results
  62. ARC-2 public roster observation
  63. ARC-3 public roster observation
  64. Public professional account profile
  65. Tufa Labs — team profiles
  66. ARC-2 public roster observation
  67. ARC-3 public roster observation
  68. The ARChitects — author and Kaggle handle credits
  69. The ARChitects — 2025 team solution
  70. The ARChitects — 2025 technical report
  71. ARC Prize — 2025 results
  72. ARC Prize — 2024 results
  73. ARC-2 public roster observation
  74. ARC-3 public roster observation
  75. ARC Prize 2025 — final Kaggle leaderboard roster
  76. ARC Prize — 2025 results
  77. IsaiahP — own Kaggle profile and direct professional identity links
  78. Tufa Labs — Isaiah Pressman biography
  79. Isaiah Pressman — Orbit Wars solution write-up
  80. ARC Prize — 2025 results
  81. ARC-3 public roster observation
  82. ARC-2 public roster observation
  83. Public Kaggle profile
  84. ARC Mutants and Team Turducken — separate dated public rosters, 19 October 2026
  85. ARC-2 public roster observation
  86. ARC-3 public roster observation
  87. Public Kaggle profile
  88. ARC Mutants and Team Turducken — separate dated public rosters, 19 October 2026
  89. ARC-2 public roster observation
  90. ARC-3 public roster observation
  91. ARC-2 public roster observation
  92. darraghdog — own public profile
  93. Darragh — direct X link from own Kaggle profile
  94. Darragh — Kaggle profile and direct professional links
  95. NVIDIA — Darragh Hanley author biography
  96. ARC-3 public roster observation
  97. ARC-2 public roster observation
  98. stefano1283 — own public profile
  99. Stefano Viel — direct X link from own Kaggle profile
  100. Stefano Viel — own Kaggle profile and direct account links
  101. Tufa Labs — Stefano Viel biography
  102. ARC-3 public roster observation
  103. ARC-2 public roster observation
  104. Own profile and direct account links
  105. Own profile and direct account links
  106. Public roster observation
  107. Own profile and direct account links
  108. Own profile and direct account links
  109. Public roster observation
  110. Public roster observation
  111. Yuvraj #2 climbs 185 places on ARC-2, from 277th to 92nd
  112. Aneesh & Yash gain 7.36 points and climb from 41st to 34th on ARC-2
  113. Ruslan Guseynov jumps from 1880th to 50th on ARC-2
  114. Giba climbs 101 places on ARC-2, from 196th to 95th
  115. DeepSeek Flash 1017 reaches 78.1% on ARC-AGI-2, with a coverage warning behind the effort curve
  116. Beam’s coding launch meets a tougher grid test: 23.5% on ARC-AGI-2
  117. Muse Spark Open reaches 57.5% on ARC-AGI-2: capable open weights, a costly top setting