论文
基于手机数据的情境感知人口位移估计:一种方法论框架
PaperDiscovery
中文标题
GeoAnalystBench:面向空间分析工作流与代码生成的地理人工智能(GeoAI)基准测试
English Title
GeoAnalystBench: A GeoAI benchmark for assessing large language models for spatial analysis workflow and code generation
未知作者
发布时间
2025/9/7 08:00:00
来源类型
journal
语言
en
摘要
中文对照
近期,GeoBenchX [20] 和 GEOBench-VLM [7] 等基准测试拓展了该领域,致力于解决地理空间任务中的多步骤、多模态及基于工具的推理挑战。ScienceAgentBench [6] 覆盖地理信息科学等四个科学学科,共包含102项任务,并基于自包含的 Python 程序进行严格的执行导向评估,使其直接适用于地理信息系统(GIS)……
English Original
Recently, benchmarks such as GeoBenchX [20] and GEOBench-VLM [7] expand this landscape by tackling multi-step, multimodal, and tool-based reasoning challenges in geospatial tasks. ScienceAgentBench [6] offers 102 tasks across four scientific disciplines, including Geographical Information Science, with rigorous execution based evaluation on self contained Python programs, making it directly relevant as a GIS ...
元数据
来源基于手机数据的情境感知人口位移估计:一种方法论框架
类型论文
抽取状态discovered
关键词
暂无关键词。