利用大型语言模型的零射击和少数射击学习能力,用于监管研究
Hamed Meshkin1, Joel Zirkle1, Ghazal Arabidarrehdor1
1Division of Applied Regulatory Science, Office of Clinical Pharmacology, Office of Translational Sciences, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, WO Bldg 64, 10903 New Hampshire Ave, Silver Spring, MD 20993, United States.
Briefings in bioinformatics
|August 23, 2024
概括
开源的大型语言模型 (LLM) 可以在本地部署用于安全的数据处理. 这些模型在提取临床药理学信息方面表现出强的表现,即使训练数据最小.
科学领域:
- 人工智能的人工智能
- 自然语言处理自然语言处理.
- 计算生物学 计算生物学
背景情况:
- 大型语言模型 (LLM) 提供先进的对话功能,但通常需要将数据传输到外部服务器.
- 在线LLM使用带来数据隐私风险,特别是敏感信息.
- 像监管机构一样,优先考虑数据保护的组织需要安全的本地AI解决方案.
研究的目的:
- 评估在安全的本地网络中实施开源LLM的可行性.
- 评估LLM在从药品标签中提取临床药理学信息方面的表现.
- 确定LLM在受监管环境中的敏感数据处理方面的有效性.
主要方法:
- 在监管机构的本地网络中实施各种开源LLM.
- 在特定的NLP任务上使用少数射击和零射击学习进行绩效评估.
- 对选定的LLM进行评估,以识别无需微调的药物暴露因素.
主要成果:
- 一些开源的LLM在最低限度的培训中实现了与传统模型相比或优于传统模型的性能.
- 一个精选的LLM准确地确定了影响药物暴露的因素,在一个大数据集上准确度为78.5%.
- 该研究证明了敏感数据分析的成功本地部署.
结论:
- 开源的LLM可以有效地在安全的本地网络中实现敏感数据任务.
- 在没有广泛的培训数据的情况下,LLM为自然语言处理提供了可行的解决方案.
- 这种方法提高了监管机构和其他优先级组织的数据隐私和安全性.
更多相关视频
相关概念视频
02:23Master Transcription Regulators
6.9K
Master transcription regulators are regulatory proteins that are predominantly responsible for regulating the expression of multiple genes. Often these genes work in concert to drive a complex process. Activation of a master transcription regulator can lead to a cascade of transcriptional activation necessary for that outcome. These regulators can directly bind to the regulatory sequences of the various genes involved, or they can indirectly regulate transcription by binding to regulatory...
6.9K
01:22Language Development
329
Children master language quickly and with relative ease, supported by both biological predisposition and reinforcement. B. F. Skinner (1957) proposed that language is learned through reinforcement, while Noam Chomsky (1965) argued that language acquisition mechanisms are biologically determined.
The critical period for language acquisition suggests that the ability to acquire language is at its peak early in life. As people age, this proficiency decreases. Language development begins very...
The critical period for language acquisition suggests that the ability to acquire language is at its peak early in life. As people age, this proficiency decreases. Language development begins very...
329
02:24Regulation of Expression Occurs at Multiple Steps
3.0K
3.0K
01:23Mechanistic Models: Compartment Models in Individual and Population Analysis
33
Mechanistic models are utilized in individual analysis using single-source data, but imperfections arise due to data collection errors, preventing perfect prediction of observed data. The mathematical equation involves known values (Xi), observed concentrations (Ci), measurement errors (εi), model parameters (ϕj), and the related function (ƒi) for i number of values. Different least-squares metrics quantify differences between predicted and observed values. The ordinary least...
33