相关实验视频

Updated: May 29, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

在临床笔记中标记实体的生成LLM的比较分析

Rodrigo Del Moral-González¹, Helena Gómez-Adorno², Orlando Ramos-Flores²

¹Posgrado en Ciencia e Ingeniería de la Computación, Universidad Nacional Autónoma de México, Circuito Escolar, Ciudad Universitaria, Coyoacán, 04510, Ciudad de México, México. rodrigodelmoral@comunidad.unam.mx.

Genomics & informatics

|February 6, 2025

概括

指令微调的大型语言模型 (LLM) 在临床命名实体识别 (NER) 任务中表现得更好. 模型在简单的输出结构中表现最好,用于识别疾病和症状.

科学领域:

自然语言处理自然语言处理.
临床信息学临床信息学
人工智能的人工智能

背景情况:

大型语言模型 (LLM) 越来越多地应用于专业领域.
零射击命名实体识别 (NER) 旨在识别没有特定领域培训数据的实体.
临床文本对自动信息提取提出了独特的挑战.

研究的目的:

评估和比较生成LLM的微调变化,以实现零射击的临床NER.
评估Llama 2和Mistral模型在确定临床实体方面的表现.
为了确定不同微调策略 (代码,聊天,指令) 对NER精度的影响.

主要方法:

利用公开可用的临床病例数据,标记疾病,症状和程序.
评估了基础,编码调整,聊天调整和指令调整的Llama 2和Mistral模型.
评估实体的识别准确性和在结构化格式中检索实体的能力.

主要成果:

在识别临床实体方面,指令微调模型的表现优于聊天微调和基准模型.
当被要求提供简单的输出结构时,模型表现出更好的性能.
具体的微调方法显著影响临床领域的零射击NER能力.

结论:

微调指令是提高临床NERLLM绩效的有希望的策略.
更简单的输出格式使LLM能够更准确地提取实体.
进一步的研究可以优化临床信息提取的LLM应用.

关键词:

这就是BLAH8的原因.临床领域是临床领域.生成型语言模型的语言模型.命名实体认可命名实体认可这是一次零射击.

更多相关视频

A Metadata Extraction Approach for Clinical Case Reports to Enable Advanced Understanding of Biomedical Concepts

A Metadata Extraction Approach for Clinical Case Reports to Enable Advanced Understanding of Biomedical Concepts

Published on: September 20, 2018

Lexical Decision Task for Studying Written Word Recognition in Adults with and without Dementia or Mild Cognitive Impairment

Lexical Decision Task for Studying Written Word Recognition in Adults with and without Dementia or Mild Cognitive Impairment

Published on: June 25, 2019

相关概念视频

Labeling Emotion

Labeling Emotion

Emotional labeling is a cognitive process that involves identifying and naming one's emotions, such as anger, fear, happiness, or sadness. It allows individuals to recognize and express their internal emotional states, a critical aspect of emotional regulation and communication. Labeling emotions requires more than mere recognition; it also involves drawing upon memory and contextual cues to understand the current situation and apply a corresponding emotional label. For instance, feeling...