Claude Opus 5 · ProgramBench(166 个保留的 golden tasks;架构与时限按协议区分) · 附件证据(ev-op5-card-programbench)

Claude Opus 5 的 ProgramBench · 166 个保留的 golden tasks;架构与时限按协议区分 官方出处。见《Claude Opus 5 System Card》System Card PDF pp.159-160, §8.9.1 ProgramBench。证据 ev-op5-card-programbench。

原文与定位

Claude Opus 5 · ProgramBench(166 个保留的 golden tasks;架构与时限按协议区分),ProgramBench · single agent配置,Hidden test pass rate 83%(具体成绩公布时间未确认)。见《Claude Opus 5 System Card》System Card PDF pp.159-160, §8.9.1 ProgramBench。 查看成绩条件

Claude Opus 5 · ProgramBench(166 个保留的 golden tasks;架构与时限按协议区分),ProgramBench · single agent配置,Hidden test pass rate 93%(具体成绩公布时间未确认)。见《Claude Opus 5 System Card》System Card PDF pp.159-160, §8.9.1 ProgramBench。 查看成绩条件

On 166 retained golden tasks, Claude Opus 5 scored 83% after episode one and 93% after episode five. Episodes continue from the previous codebase with a fresh context budget up to 1M tokens.
官方材料
Claude Opus 5 System Card · PDF 第 160 页
定位
附件 · System Card PDF pp.159-160, §8.9.1 ProgramBench
材料版本
Current saved PDF/TXT: 本地留存副本 and current-opus-5-system-card.txt; cover dated 2026-07-24; p.2 changelog records an update on 2026-08-19; later revision date, if any, unknown.
版本公布
未确认
披露时间
未确认 · 首发/后补未确认
采集时间
历史依据
暂无可靠历史版本依据
引用文本

Claude Opus 5 · ProgramBench(166 个保留的 golden tasks;架构与时限按协议区分),ProgramBench · single agent配置,Hidden test pass rate 83%(具体成绩公布时间未确认);Claude Opus 5 · ProgramBench(166 个保留的 golden tasks;架构与时限按协议区分),ProgramBench · single agent配置,Hidden test pass rate 93%(具体成绩公布时间未确认)。官方出处 Claude Opus 5 System Card(附件,Current saved PDF/TXT: 本地留存副本 and current-opus-5-system-card.txt; cover dated 2026-07-24; p.2 changelog records an update on 2026-08-19; later revision date, if any, unknown.)System Card PDF pp.159-160, §8.9.1 ProgramBench,https://www.anthropic.com/claude-opus-5-system-card#page=160。Benchlog https://benchlog.lekshome.cc/evidence/claude-opus-5/ev-op5-card-programbench/

关联成绩与条件

Claude Opus 5 · ProgramBench 成绩与配置
指标与成绩配置与评测条件证据关系
Hidden test pass rate83%原文 83% · 单位 %
公布 未确认
ProgramBench · single agent
设置
Card-wide default max effort; one agent; 166 retained golden tasks; five sequential episodes, each with a fresh context budget up to 1M tokens; no internet or decompilation tools.
未知项
未另记
工具
工具条件未记录
推理预算
max
框架
未记录
协议
ProgramBench; 166 golden tasks; hidden-test pass rate after episode 1.
条件
Single-agent; episode 1; each episode has a fresh context budget up to 1M tokens; max effort by card-wide default.

直接支持该成绩

在排名中查看此成绩
Hidden test pass rate93%原文 93% · 单位 %
公布 未确认
ProgramBench · single agent
设置
Card-wide default max effort; one agent; 166 retained golden tasks; five sequential episodes, each with a fresh context budget up to 1M tokens; no internet or decompilation tools.
未知项
未另记
工具
工具条件未记录
推理预算
max
框架
未记录
协议
ProgramBench; 166 golden tasks; hidden-test pass rate after episode 5.
条件
Single-agent; five sequential episodes continue from the previous codebase; fresh context budget up to 1M tokens per episode; max effort by card-wide default.

直接支持该成绩

在排名中查看此成绩