Claude Haiku 5.5 · OSWorld · 2.1, 82-task offline no-internet subset, partial score · 成绩与配置

Claude Haiku 5.5 的 OSWorld · 2.1, 82-task offline no-internet subset, partial score 记录。本次发布有 2 条成绩记录,含 2 条精确值(可含历史记录),附 2 条来源定位。

成绩与评测条件

Claude Haiku 5.5 · OSWorld 成绩与配置
指标与成绩配置与评测条件出处
Partial score72.4%原文 72.4% partial (offline subset) · 单位 %
公布 未确认
OSWorld 2.1 offline subset
设置
OSWorld 2.1 official offline subset: 82 of 108 tasks, no internet, 1080p, max 500 action steps; partial score and strict pass rate are pass@1 averaged over five independent attempts per task at max effort.
未知项
System Card 当前修订日期未给出。
工具
工具条件未记录
推理预算
max
框架
未记录
协议
OSWorld 2.1 official offline subset: 82/108 tasks; no internet; 1080p; maximum 500 action steps; pass@1 average across five attempts per task, max effort.
条件
Partial score is mean per-task credit across checkpoints.
在排名中查看此成绩
Strict pass rate37.1%原文 37.1% strict pass (offline subset) · 单位 %
公布 未确认
OSWorld 2.1 offline subset
设置
OSWorld 2.1 official offline subset: 82 of 108 tasks, no internet, 1080p, max 500 action steps; partial score and strict pass rate are pass@1 averaged over five independent attempts per task at max effort.
未知项
System Card 当前修订日期未给出。
工具
工具条件未记录
推理预算
max
框架
未记录
协议
OSWorld 2.1 official offline subset: 82/108 tasks; no internet; 1080p; maximum 500 action steps; pass@1 average across five attempts per task, max effort.
条件
Strict pass rate counts tasks where every checkpoint is satisfied.
在排名中查看此成绩

出处与披露

测试版本与指标

别名 OSWorld-2.1 offline subset

  • Partial score · % · 越高越好
  • Strict pass rate · % · 越高越好