GPT-5.5 · GeneBench(2026-04-23 report; 103 evaluations across 10 domains) · 附件证据(ev-gpt55-0065)

GPT-5.5 的 GeneBench · 2026-04-23 report; 103 evaluations across 10 domains 官方出处。见《GeneBench: Assessing AI Agents for Multi-Stage Inference Problems in Genomics and Quantitative Biology》GeneBench report / Abstract and Introduction。证据 ev-gpt55-0065。

原文与定位

GeneBench contains 103 evaluations across 10 domains. Each task is a multi-stage quantitative analysis; overall pass rate is an unweighted mean of per-problem pass rates across repeated runs.
官方材料
GeneBench: Assessing AI Agents for Multi-Stage Inference Problems in Genomics and Quantitative Biology · PDF 第 8 页
定位
附件 · GeneBench report / Abstract and Introduction
材料版本
31-page report; 103 evaluations across 10 domains; table includes full-suite pass rates by reasoning effort.
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

GPT-5.5 · GeneBench(2026-04-23 report; 103 evaluations across 10 domains),披露记录。官方出处 GeneBench: Assessing AI Agents for Multi-Stage Inference Problems in Genomics and Quantitative Biology(附件,31-page report; 103 evaluations across 10 domains; table includes full-suite pass rates by reasoning effort.)GeneBench report / Abstract and Introduction,https://cdn.openai.com/pdf/6dc7175d-d9e7-4b8d-96b8-48fe5798cd5b/oai_genebench_benchmark.pdf#page=8。Benchlog https://benchlog.lekshome.cc/evidence/gpt-5-5/ev-gpt55-0065/

关联成绩与条件

此摘录未关联到已收录的成绩记录。