• Home
  • Services
  • About
  • Contact
  • Blog
  • 知財活動のROICへの貢献
  • 生成AIを活用した知財戦略の策定方法
  • 生成AIとの「壁打ち」で、新たな発明を創出する方法

​
​よろず知財コンサルティングのブログ

Johan Land氏ARC-AGI-2で72.9%達成

4/2/2026

0 Comments

 
2026年2月3日、ARC-AGI-2 Leaderboardに、 Johan Land氏がGPT-5.2に手を加えたシステムが、ARC-AGI-2(Abstraction and Reasoning Corpus for AGI - Version 2)において72.9%という驚異的な正答率を記録したことが正式に記載されました。
ARC-AGI-2は、人間にとっては簡単でも、現在のAIシステムにとっては非常に難しいタスクで構成された抽象的推論能力(蓄積された知識ではなく、新しい状況で推論し、問題を解決する能力=流動的知性)を測定するベンチマークの新しいバージョンで、この記録は、直前のState-of-the-Art(SOTA)であったGPT-5.2のスコア(54.2%)を大幅に上回るものであり、AIが特定領域のスキルだけでなく、未知の問題に対する適応能力において人間レベル(平均的な人間は約60-66%とされる)を超えたことを示しています。
このARC-AGI-2で72.9%達成という情報について、生成AIに深掘りさせました。さらに、結果をNotebookLMでインフォグラフィック、スライド資料にさせました。
なお、生成AIによる調査・分析結果は、公開された情報からだけの分析であり、必ずしも実情を示したものではないこと、誤った情報も含まれていることについてはご留意されたうえで、ご参照ください。
 
ARC-AGI-2 Leaderboard
https://arcprize.org/leaderboard
 
 
Johan Land Achieves 72.9% on ARC-AGI-2
On February 3, 2026, the ARC-AGI-2 Leaderboard officially recorded that a system developed by Johan Land, based on a modified version of GPT-5.2, achieved an astonishing accuracy of 72.9% on ARC-AGI-2 (Abstraction and Reasoning Corpus for AGI – Version 2).
ARC-AGI-2 is a new version of a benchmark designed to measure abstract reasoning ability—that is, fluid intelligence, the capacity to reason and solve problems in novel situations rather than relying on accumulated knowledge. While the tasks are relatively easy for humans, they are extremely challenging for current AI systems. This result significantly surpasses the previous state-of-the-art score of 54.2%, achieved by GPT-5.2 itself, and suggests that AI has exceeded average human-level performance in adaptability to unfamiliar problems (with average human scores generally estimated at around 60–66%), not merely in narrow, domain-specific skills.
I conducted an in-depth analysis of this 72.9% ARC-AGI-2 result using generative AI, and further converted the findings into infographics and slide materials using NotebookLM.
Please note that the investigations and analyses conducted by generative AI are based solely on publicly available information and may not fully reflect real-world conditions. They may also contain inaccuracies, so readers are advised to consult the results with appropriate caution.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
0 Comments



Leave a Reply.

    著者

    萬秀憲

    アーカイブ

    January 2026
    December 2025
    November 2025
    October 2025
    September 2025
    August 2025
    July 2025
    June 2025
    May 2025
    April 2025
    March 2025
    February 2025
    January 2025
    December 2024
    November 2024
    October 2024
    September 2024
    August 2024
    July 2024
    June 2024
    May 2024
    April 2024
    March 2024
    February 2024
    January 2024
    December 2023
    November 2023
    October 2023
    September 2023
    August 2023
    July 2023
    June 2023
    May 2023
    April 2023
    March 2023
    February 2023
    January 2023
    December 2022
    November 2022
    October 2022
    September 2022
    August 2022
    July 2022
    June 2022
    May 2022
    April 2022
    March 2022
    February 2022
    January 2022
    December 2021
    November 2021
    October 2021
    September 2021
    August 2021
    July 2021
    June 2021
    May 2021
    April 2021
    March 2021
    February 2021
    January 2021
    December 2020
    November 2020
    October 2020
    September 2020
    August 2020
    July 2020
    June 2020

    カテゴリー

    All

    RSS Feed

Copyright © よろず知財戦略コンサルティング All Rights Reserved.
サイトはWeeblyにより提供され、お名前.comにより管理されています
  • Home
  • Services
  • About
  • Contact
  • Blog
  • 知財活動のROICへの貢献
  • 生成AIを活用した知財戦略の策定方法
  • 生成AIとの「壁打ち」で、新たな発明を創出する方法