Event ·Official first-hand

大模型与本体排序器融合提升罕见病诊断

该研究提出了一种行为融合模型,结合大语言模型和本体排序器进行罕见病诊断。在Phenopacket Store和RAMEDIS数据集上,融合方法将Phenomizer Recall@1分别提升7.86和20.18个百分点。当与DeepSeek-V4-Flash API结合时,融合模型将Recall@1从0.1657提升至0.2176,提升5.19个百分点,无需重新训练。90.8%的正确融合诊断保留了可检查的本体证据。

Current read

研究将LLM与本体排序器融合,提升罕见病诊断准确率,同时保留可追溯的证据链。

4 sources73° AI heatLast updated 2026/9/2 11:44:21

Evidence

4 sources
Primary sourcearXiv: DeepSeek
View source
Related source 1shao__meng
View 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.