Claude Opus 5.5 · AA-Briefcase · v1.1 · 成绩与配置

Claude Opus 5.5 的 AA-Briefcase · v1.1 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 1 条来源定位。

成绩与评测条件

Claude Opus 5.5 · AA-Briefcase 成绩与配置
指标与成绩配置与评测条件出处
Elo rating1705 Elo原文 1705 Elo · 单位 Elo
公布 未确认
AA-Briefcase v1.1 high effort
设置
Explicit high effort; Artificial Analysis independent harness.
未知项
Run date is not stated.
工具
工具条件未记录
推理预算
high
框架
Artificial Analysis independent harness
协议
AA-Briefcase v1.1; Artificial Analysis independent harness; high effort.
条件
Long-horizon project tasks; score is not a public-task success percentage. Source-wide conditions: AA-Briefcase v1.1; Artificial Analysis independent evaluation; multi-week linked projects with thousands of input source files.
在排名中查看此成绩
Elo rating1780 Elo原文 1780 Elo · 单位 Elo
公布 未确认
AA-Briefcase v1.1 xhigh effort
设置
Explicit xhigh effort; Artificial Analysis independent harness.
未知项
Run date is not stated.
工具
工具条件未记录
推理预算
xhigh
框架
Artificial Analysis independent harness
协议
AA-Briefcase v1.1; Artificial Analysis independent harness; xhigh effort.
条件
About 41% fewer output tokens than max is reported; no token budget is inferred. Source-wide conditions: AA-Briefcase v1.1; Artificial Analysis independent evaluation; multi-week linked projects with thousands of input source files.
在排名中查看此成绩
Elo rating1822 Elo原文 1822 Elo · 单位 Elo
公布 未确认
AA-Briefcase v1.1 max effort
设置
Explicit max effort; Artificial Analysis independent harness.
未知项
Run date is not stated.
工具
工具条件未记录
推理预算
max
框架
Artificial Analysis independent harness
协议
AA-Briefcase v1.1; Artificial Analysis independent harness; max effort.
条件
Distinct from GDPval-AA v2.1; multi-week linked project tasks. Source-wide conditions: AA-Briefcase v1.1; Artificial Analysis independent evaluation; multi-week linked projects with thousands of input source files.
在排名中查看此成绩

出处与披露

测试版本与指标

别名 AA-Briefcase v1.1

  • Elo rating · Elo · 越高越好
  • Rubric score · % · 越高越好