使用调整指令的大型语言模型来识别警方事件叙述中的脆弱性指标.
Sam Relins1,2,3, Daniel Birks2,3, Charlie Lloyd1,3
1School for Business and Society, University of York, York, UK.
概括
大型语言模型 (LLM) 可以有效地为弱势群体编码警察报告,显示与人类编码器的高度一致. 这项技术为分析叙事数据提供了一种资源高效的方法.
科学领域:
- 计算社会科学 计算社会科学
- 犯罪学中的人工智能
背景情况:
- 警方叙事报告提供了与弱势群体互动的见解.
- 这些报告的定性编码是资源密集的.
研究的目的:
- 评估大型语言模型 (LLM) 用于复制警察叙事报告的人类定性编码.
- 评估LLM在识别特定弱点方面的有效性:精神疾病健康,物质滥用,酒精依赖和无家可归.
主要方法:
- 在波士顿警察局的叙事报告上,人类生成的标签和LLM生成的标签的比较.
- 评估各种LLM大小和提示策略.
- 对标签变异性的分析和用于偏见检测的反事实实验 (性别,种族).
主要成果:
- LLM与人类编码者达成高度一致,特别是对于非易受攻击的病例.
- 较大的模型和量身定制的提示改善了人-LLM协议,根据漏洞类型的变化.
- 反事实分析揭示了LLM分类中与性别和种族相关的最小偏见.
结论:
- 简单的LLM可以显著减少对叙事数据集的手动编码要求.
- 人与LLM的协作方法提高了编码的特异性,透明度和可复制性.
- 法学士申请为犯罪学和相关领域提供了机会.
相关概念视频
Steps in Outbreak Investigation
476
In the ever-evolving field of public health, statistical analysis serves as a cornerstone for understanding and managing disease outbreaks. By leveraging various statistical tools, health professionals can predict potential outbreaks, analyze ongoing situations, and devise effective responses to mitigate impact. For that to happen, there are a few possible stages of the analysis:
476
Types of Reports II: Incident or Occurrence Report
1.2K
An Incident or Occurrence Report in a healthcare setting is a crucial document used to record any unexpected occurrence that may or may not have affected a patient, employee, or visitor. Such reports are critical to improving patient safety and include all details leading up to and including the event.
Purposes:
In the healthcare industry, reports play a crucial role in documenting incidents within an agency. The primary objective of these reports is to ensure patient safety, uphold the...
Purposes:
In the healthcare industry, reports play a crucial role in documenting incidents within an agency. The primary objective of these reports is to ensure patient safety, uphold the...
1.2K

