Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts) · 附件证据(ev-h55-card-physicianbench)
Claude Haiku 5.5 的 PhysicianBench · 100 physician tasks in an electronic health record; pass rate across attempts 官方出处。见《Claude Haiku 5.5 System Card》System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A。证据 ev-h55-card-physicianbench。
Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / low配置,Pass rate 17.8%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / medium配置,Pass rate 25.2%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / high配置,Pass rate 31.6%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / xhigh配置,Pass rate 35.8%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / max配置,Pass rate 43%(具体成绩公布时间未确认)。见《Claude Haiku 5.5 System Card》System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A。 查看成绩条件
Claude Haiku 5.5 passed 43.0% of PhysicianBench attempts at max; the figure labels all effort-level scores.
附件 · System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.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 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / low配置,Pass rate 17.8%(具体成绩公布时间未确认);Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / medium配置,Pass rate 25.2%(具体成绩公布时间未确认);Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / high配置,Pass rate 31.6%(具体成绩公布时间未确认);Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / xhigh配置,Pass rate 35.8%(具体成绩公布时间未确认);Claude Haiku 5.5 · PhysicianBench(100 physician tasks in an electronic health record; pass rate across attempts),Healthcare evaluations / max配置,Pass rate 43%(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.132-134, §8.11.2 and Figure 8.11.3.A,https://www.anthropic.com/claude-haiku-5-5-system-card#page=134。Benchlog https://benchlog.lekshome.cc/evidence/claude-haiku-5-5/ev-h55-card-physicianbench/
关联成绩与条件
Claude Haiku 5.5 · PhysicianBench 成绩与配置
指标与成绩
配置与评测条件
证据关系
Pass rate17.8%原文 17.8% · 单位 %
公布 未确认
Healthcare evaluations / low
设置
Safety classifiers enabled; low effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
工具条件未记录
推理预算
low
框架
未记录
协议
PhysicianBench; five attempts per task at each effort level.
条件
low effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; medium effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
工具条件未记录
推理预算
medium
框架
未记录
协议
PhysicianBench; five attempts per task at each effort level.
条件
medium effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; high effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
工具条件未记录
推理预算
high
框架
未记录
协议
PhysicianBench; five attempts per task at each effort level.
条件
high effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; xhigh effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
工具条件未记录
推理预算
xhigh
框架
未记录
协议
PhysicianBench; five attempts per task at each effort level.
条件
xhigh effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; max effort; run counts, scoring variants, and dates are recorded per result.
未知项
Healthcare evaluations use benchmark-specific run counts and scoring variants.
工具
工具条件未记录
推理预算
max
框架
未记录
协议
PhysicianBench; five attempts per task at each effort level.
条件
max effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts) · Attachment evidence (ev-h55-card-physicianbench)
Official source for PhysicianBench · 100 physician tasks in an electronic health record; pass rate across attempts in Claude Haiku 5.5. See “Claude Haiku 5.5 System Card” at System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A. Evidence ev-h55-card-physicianbench.
Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / low, Pass rate 17.8% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / medium, Pass rate 25.2% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / high, Pass rate 31.6% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / xhigh, Pass rate 35.8% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / max, Pass rate 43% (exact score publication date unconfirmed). See “Claude Haiku 5.5 System Card” at System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A. View score conditions
Claude Haiku 5.5 passed 43.0% of PhysicianBench attempts at max; the figure labels all effort-level scores.
Attachment · System Card PDF pp.132-134, §8.11.2 and Figure 8.11.3.A
Material version
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。
Version published
Disclosure date
Unconfirmed · Launch or later disclosure unconfirmed
Collected
Historical basis
No reliable historical version evidence available
Citation text
Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / low, Pass rate 17.8% (exact score publication date unconfirmed); Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / medium, Pass rate 25.2% (exact score publication date unconfirmed); Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / high, Pass rate 31.6% (exact score publication date unconfirmed); Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / xhigh, Pass rate 35.8% (exact score publication date unconfirmed); Claude Haiku 5.5 · PhysicianBench (100 physician tasks in an electronic health record; pass rate across attempts), configuration Healthcare evaluations / max, Pass rate 43% (exact score publication date unconfirmed). Official source: Claude Haiku 5.5 System Card (Attachment, 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.132-134, §8.11.2 and Figure 8.11.3.A, https://www.anthropic.com/claude-haiku-5-5-system-card#page=134. Benchlog https://benchlog.lekshome.cc/evidence/claude-haiku-5-5/ev-h55-card-physicianbench/?lang=en
Related scores and conditions
Claude Haiku 5.5 · PhysicianBench scores and configurations
Metric and score
Configuration and evaluation conditions
Evidence relation
Pass rate17.8%Reported 17.8% · Unit %
Published Unconfirmed
Healthcare evaluations / low
Settings
Safety classifiers enabled; low effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tool access not recorded
Reasoning budget
low
Framework
not recorded
Protocol
PhysicianBench; five attempts per task at each effort level.
Conditions
low effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; medium effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tool access not recorded
Reasoning budget
medium
Framework
not recorded
Protocol
PhysicianBench; five attempts per task at each effort level.
Conditions
medium effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; high effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tool access not recorded
Reasoning budget
high
Framework
not recorded
Protocol
PhysicianBench; five attempts per task at each effort level.
Conditions
high effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; xhigh effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tool access not recorded
Reasoning budget
xhigh
Framework
not recorded
Protocol
PhysicianBench; five attempts per task at each effort level.
Conditions
xhigh effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.
Safety classifiers enabled; max effort; run counts, scoring variants, and dates are recorded per result.
Unknowns
Healthcare evaluations use benchmark-specific run counts and scoring variants.
Tools
Tool access not recorded
Reasoning budget
max
Framework
not recorded
Protocol
PhysicianBench; five attempts per task at each effort level.
Conditions
max effort; safety classifiers enabled. Low through xhigh runs were made October 5; max runs were made October 1-2. Max score was also reported in §8.11.2. Labeled figure value.