Event ·Official first-hand

RAME框架在谷物育种信息提取任务中夺冠

研究人员提出RAME框架，在CCL2026-Eval Task 5任务中排名第一。该框架基于DeepSeek-V4-Flash模型，结合检索增强和三种提示词策略。RAME在NER任务得分为0.730，RE任务得分为0.346，总分0.499，超越GPT-5.5基线11.4%。

Current read

DeepSeek团队用RAME框架在谷物育种信息提取任务击败GPT-5.5，代码已开源。

4 sources58° AI heatLast updated 2026/9/7 07:30:49

Evidence

4 sources

Primary sourcearXiv: DeepSeek

2026/9/7 07:30:49View source

Related source 1ollama

2026/9/5 23:04:45View source

Related source 2IT之家

2026/9/8 22:39:58View source

Related source 3Geek

2026/9/9 02:31:37View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story