Claude Haiku 5.5 · WANDR(500-task wide research benchmark) · 附件证据(ev-h55-card-wandr)

Claude Haiku 5.5 的 WANDR · 500-task wide research benchmark 官方出处。见《Claude Haiku 5.5 System Card》System Card PDF pp.122-123, Figure 8.8.3.A。证据 ev-h55-card-wandr。

原文与定位

Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / low配置,Soft F1 3.5%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.122-123, Figure 8.8.3.A。 查看成绩条件

Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / medium配置,Soft F1 12.9%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.122-123, Figure 8.8.3.A。 查看成绩条件

Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / high配置,Soft F1 37.3%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.122-123, Figure 8.8.3.A。 查看成绩条件

Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / xhigh配置,Soft F1 45.9%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.122-123, Figure 8.8.3.A。 查看成绩条件

Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / max配置,Soft F1 49.9%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.122-123, Figure 8.8.3.A。 查看成绩条件

WANDR labels Haiku 5.5 at 3.5%, 12.9%, 37.3%, 45.9%, and 49.9% for low through max effort.
官方材料
Claude Haiku 5.5 System Card · PDF 第 123 页
定位
附件 · System Card PDF pp.122-123, Figure 8.8.3.A
材料版本
144 页 PDF/TXT:本地留存副本、current-haiku-5-5-system-card.txt;封面日期 2026-10-07;正文 pp.1-144 已全文核读,全部63个含图/向量图表页实际视觉检查(本轮补看53页,合并此前10页),无未读页或未查看图表页;此为 current 卡片,不作为首发日附件证据。
版本公布
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 Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / low配置,Soft F1 3.5%(具体成绩公布时间未确认);Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / medium配置,Soft F1 12.9%(具体成绩公布时间未确认);Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / high配置,Soft F1 37.3%(具体成绩公布时间未确认);Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / xhigh配置,Soft F1 45.9%(具体成绩公布时间未确认);Claude Haiku 5.5 · WANDR(500-task wide research benchmark),System Card effort / max配置,Soft F1 49.9%(具体成绩公布时间未确认)。官方出处 Claude Haiku 5.5 System Card(附件,144 页 PDF/TXT:本地留存副本、current-haiku-5-5-system-card.txt;封面日期 2026-10-07;正文 pp.1-144 已全文核读,全部63个含图/向量图表页实际视觉检查(本轮补看53页,合并此前10页),无未读页或未查看图表页;此为 current 卡片,不作为首发日附件证据。)System Card PDF pp.122-123, Figure 8.8.3.A,https://www.anthropic.com/claude-haiku-5-5-system-card#page=123。Benchlog https://benchlog.lekshome.cc/evidence/claude-haiku-5-5/ev-h55-card-wandr/

关联成绩与条件

Claude Haiku 5.5 · WANDR 成绩与配置
指标与成绩配置与评测条件证据关系
Soft F13.5%原文 3.5% · 单位 %
公布 未确认
System Card effort / low
设置
Adaptive thinking; low effort, explicitly labeled in the source; benchmark-specific protocol is preserved in each result's conditions.
未知项
Sampling and trial counts vary by evaluation and are recorded per result.
工具
允许工具
推理预算
low
框架
未记录
协议
WANDR; 500 wide-research tasks; web search, web fetch, programmatic tool calling, code execution; 980k-token budget; soft F1 averaged across tasks.
条件
low effort. Anthropic used a frozen offline index and Claude Opus 4.6 judge; tool/grading differences mean results are not directly comparable to Perplexity’s. Labeled figure value.

直接支持该成绩

在排名中查看此成绩
Soft F112.9%原文 12.9% · 单位 %
公布 未确认
System Card effort / medium
设置
Adaptive thinking; medium effort, explicitly labeled in the source; benchmark-specific protocol is preserved in each result's conditions.
未知项
Sampling and trial counts vary by evaluation and are recorded per result.
工具
允许工具
推理预算
medium
框架
未记录
协议
WANDR; 500 wide-research tasks; web search, web fetch, programmatic tool calling, code execution; 980k-token budget; soft F1 averaged across tasks.
条件
medium effort. Anthropic used a frozen offline index and Claude Opus 4.6 judge; tool/grading differences mean results are not directly comparable to Perplexity’s. Labeled figure value.

直接支持该成绩

在排名中查看此成绩
Soft F137.3%原文 37.3% · 单位 %
公布 未确认
System Card effort / high
设置
Adaptive thinking; high effort, explicitly labeled in the source; benchmark-specific protocol is preserved in each result's conditions.
未知项
Sampling and trial counts vary by evaluation and are recorded per result.
工具
允许工具
推理预算
high
框架
未记录
协议
WANDR; 500 wide-research tasks; web search, web fetch, programmatic tool calling, code execution; 980k-token budget; soft F1 averaged across tasks.
条件
high effort. Anthropic used a frozen offline index and Claude Opus 4.6 judge; tool/grading differences mean results are not directly comparable to Perplexity’s. Labeled figure value.

直接支持该成绩

在排名中查看此成绩
Soft F145.9%原文 45.9% · 单位 %
公布 未确认
System Card effort / xhigh
设置
Adaptive thinking; xhigh effort, explicitly labeled in the source; benchmark-specific protocol is preserved in each result's conditions.
未知项
Sampling and trial counts vary by evaluation and are recorded per result.
工具
允许工具
推理预算
xhigh
框架
未记录
协议
WANDR; 500 wide-research tasks; web search, web fetch, programmatic tool calling, code execution; 980k-token budget; soft F1 averaged across tasks.
条件
xhigh effort. Anthropic used a frozen offline index and Claude Opus 4.6 judge; tool/grading differences mean results are not directly comparable to Perplexity’s. Labeled figure value.

直接支持该成绩

在排名中查看此成绩
Soft F149.9%原文 49.9% · 单位 %
公布 未确认
System Card effort / max
设置
Adaptive thinking; max effort, explicitly labeled in the source; benchmark-specific protocol is preserved in each result's conditions.
未知项
Sampling and trial counts vary by evaluation and are recorded per result.
工具
允许工具
推理预算
max
框架
未记录
协议
WANDR; 500 wide-research tasks; web search, web fetch, programmatic tool calling, code execution; 980k-token budget; soft F1 averaged across tasks.
条件
max effort. Anthropic used a frozen offline index and Claude Opus 4.6 judge; tool/grading differences mean results are not directly comparable to Perplexity’s. Labeled figure value.

直接支持该成绩

在排名中查看此成绩