Claude Opus 4.5 · LAB-Bench FigQA(FutureHouse FigQA; prompting and tools are recorded per result) · 附件证据(ev-op45-card-figqa)

Claude Opus 4.5 的 LAB-Bench FigQA · FutureHouse FigQA; prompting and tools are recorded per result 官方出处。见《Claude Opus 4.5 System Card》PDF pp.33-34, §2.21, Figure 2.21.A。证据 ev-op45-card-figqa。

原文与定位

Claude Opus 4.5 · LAB-Bench FigQA(FutureHouse FigQA; prompting and tools are recorded per result),LAB-Bench FigQA, no tools配置,Score 54.9%(具体成绩公布时间未确认)。见《Claude Opus 4.5 System Card》PDF pp.33-34, §2.21, Figure 2.21.A。 查看成绩条件

Claude Opus 4.5 · LAB-Bench FigQA(FutureHouse FigQA; prompting and tools are recorded per result),LAB-Bench FigQA, image cropping tool配置,Score 69.2%(具体成绩公布时间未确认)。见《Claude Opus 4.5 System Card》PDF pp.33-34, §2.21, Figure 2.21.A。 查看成绩条件

Without tools and with extended thinking mode off ... 54.9%. With a simple image cropping tool and a reasoning token budget of 32,768 tokens ... 69.2%.
官方材料
Claude Opus 4.5 System Card · PDF 第 34 页
定位
附件 · PDF pp.33-34, §2.21, Figure 2.21.A
材料版本
Current saved PDF/TXT: 本地留存副本 and .txt; cover November 2025, change log 2025-11-24; current binary is not established as launch-day bytes; fully reviewed 2026-11-16.
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 4.5 · LAB-Bench FigQA(FutureHouse FigQA; prompting and tools are recorded per result),LAB-Bench FigQA, no tools配置,Score 54.9%(具体成绩公布时间未确认);Claude Opus 4.5 · LAB-Bench FigQA(FutureHouse FigQA; prompting and tools are recorded per result),LAB-Bench FigQA, image cropping tool配置,Score 69.2%(具体成绩公布时间未确认)。官方出处 Claude Opus 4.5 System Card(附件,Current saved PDF/TXT: 本地留存副本 and .txt; cover November 2025, change log 2025-11-24; current binary is not established as launch-day bytes; fully reviewed 2026-11-16.)PDF pp.33-34, §2.21, Figure 2.21.A,https://www.anthropic.com/claude-opus-4-5-system-card#page=34。Benchlog https://benchlog.lekshome.cc/evidence/claude-opus-4-5/ev-op45-card-figqa/

关联成绩与条件

Claude Opus 4.5 · LAB-Bench FigQA 成绩与配置
指标与成绩配置与评测条件证据关系
Score54.9%原文 54.9% · 单位 %
公布 未确认
LAB-Bench FigQA, no tools
设置
Extended thinking off; no tools; zero-shot prompting.
未知项
Trial count is not stated.
工具
禁止工具
推理预算
未记录
框架
未记录
协议
LAB-Bench FigQA; zero-shot prompting; 95% confidence interval shown in the figure.
条件
No tools; extended thinking off.

直接支持该成绩

在排名中查看此成绩
Score69.2%原文 69.2% · 单位 %
公布 未确认
LAB-Bench FigQA, image cropping tool
设置
Simple image cropping tool; 32,768 reasoning-token budget; zero-shot prompting.
未知项
Trial count is not stated.
工具
允许工具
推理预算
思考预算 32,768
框架
未记录
协议
LAB-Bench FigQA; zero-shot prompting; 95% confidence interval shown in the figure.
条件
Simple image cropping tool; 32,768 reasoning-token budget.

直接支持该成绩

在排名中查看此成绩