GPT-6 Astra · ARC-AGI-3 · Responses API harness with two settings changed to better match real-world performance · 成绩与配置

GPT-6 Astra 的 ARC-AGI-3 · Responses API harness with two settings changed to better match real-world performance 记录。本次发布有 3 条成绩记录,含 3 条精确值(可含历史记录),附 5 条来源定位。

成绩与评测条件

GPT-6 Astra · ARC-AGI-3 成绩与配置
指标与成绩配置与评测条件出处
Score99.9%原文 99.9% · 单位 %
公布 未确认
GPT-6 Astra table result
设置
The article reports the maximum score across efforts; exact effort is not specified for this row.
未知项
Do not infer Max or another specific effort from the table alone.
工具
工具条件未记录
推理预算
未记录
框架
responses API harness
协议
未说明
条件
GPT-6 Astra; table reports the maximum score across efforts; the specific effort is not disclosed unless an exact chart result identifies it. Responses API harness; two settings changed to better match real-world performance. ARC-AGI-3 使用 Responses API harness,两项设置调整用于贴近真实表现,未专门针对 ARC-AGI-3。
在排名中查看此成绩
Score100%原文 100% · 单位 %
公布 未确认
GPT-6 Astra, configuration unspecified
设置
The official point label or tooltip does not disclose reasoning effort or other configuration.
未知项
Do not infer effort from point order or chart coordinates.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
未说明
条件
GPT-6 Astra; reasoning effort only where directly labeled.
在排名中查看此成绩
Levels exceeding human action-efficiency baseline96%原文 96% · 单位 %
公布 未确认
GPT-6 Astra table result
设置
The article reports the maximum score across efforts; exact effort is not specified for this row.
未知项
Do not infer Max or another specific effort from the table alone.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
ARC-AGI-3 human action-efficiency comparison
条件
The quoted percentage is the share of levels on which Astra exceeded the human action-efficiency baseline; not the benchmark score.
在排名中查看此成绩

出处与披露

测试版本与指标

  • Score · % · 越高越好
  • Levels exceeding human action-efficiency baseline · % · 越高越好