Gemini 3.1 Pro · SciCode(Scientific research coding) 59% · 附件证据(eval-scicode)

Gemini 3.1 Pro 的 SciCode · Scientific research coding 官方出处。Gemini 3.1 Pro Thinking (High) · 59%。见《Gemini 3.1 Pro Model Evaluation - Approach, Methodology & Results》PDF p.3, results table, SciCode row, Gemini 3.1 Pro Thinking (High) column。证据 eval-scicode。

原文与定位

Gemini 3.1 Pro · SciCode(Scientific research coding),Gemini 3.1 Pro Thinking (High)配置,Accuracy 59%(具体成绩公布时间未确认)。见《Gemini 3.1 Pro Model Evaluation - Approach, Methodology & Results》PDF p.3, results table, SciCode row, Gemini 3.1 Pro Thinking (High) column。 查看成绩条件

SciCode / Gemini 3.1 Pro Thinking (High): 59%
官方材料
Gemini 3.1 Pro Model Evaluation - Approach, Methodology & Results · PDF 第 3 页
定位
附件 · PDF p.3, results table, SciCode row, Gemini 3.1 Pro Thinking (High) column
材料版本
3-page evaluation methodology PDF; results as of February 2026; PDF states Published: February, 2026
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

Gemini 3.1 Pro · SciCode(Scientific research coding),Gemini 3.1 Pro Thinking (High)配置,Accuracy 59%(具体成绩公布时间未确认)。官方出处 Gemini 3.1 Pro Model Evaluation - Approach, Methodology & Results(附件,3-page evaluation methodology PDF; results as of February 2026; PDF states Published: February, 2026)PDF p.3, results table, SciCode row, Gemini 3.1 Pro Thinking (High) column,https://deepmind.google/models/evals-methodology/gemini-3-1-pro#page=3。Benchlog https://benchlog.lekshome.cc/evidence/gemini-3-1-pro/eval-scicode/

关联成绩与条件

Gemini 3.1 Pro · SciCode 成绩与配置
指标与成绩配置与评测条件证据关系
Accuracy59%原文 59% · 单位 %
公布 未确认
Gemini 3.1 Pro Thinking (High)
设置
Gemini API model id gemini-3.1-pro-preview; default sampling settings; evaluation table column specifies Thinking (High). All scores are pass@1 except where noted.
未知项
Exact thinking budget and sampling parameters are not listed.
工具
工具条件未记录
推理预算
high
框架
未记录
协议
Result sourced from Artificial Analysis.
条件
Scientific research coding.

直接支持该成绩

在排名中查看此成绩