jacob-valdez/synthux-economy-r3-w016
收藏资源简介:
SynthUX计算机使用数据集(经济模拟r3)是一个由SynthUX生成的基于视觉的计算机使用轨迹数据集。每个记录代表一个工人在模拟公司内的设备会话:一个目标被扩展为节点树,终端节点通过低级别鼠标/键盘输入驱动真实桌面模拟器应用程序(如终端、笔记、VS Code、浏览器、Slack/Teams、邮件、表格、幻灯片等),观察到的轨迹被记录为屏幕视频和每个节点的帧。该数据集支持应用程序原生多应用程序执行(无工作台覆盖)、每个参与者内容多样性、跨操作系统消息传递以及大型多公司经济模拟。每个经济模拟作为独立的数据集仓库发布。数据集包含数据文件(data.jsonl)、每个终端节点的截图(screenshots/)和完整屏幕录制视频(videos/)。记录模式包括轨迹ID、参与者信息、环境、目标、节点树、输入事件、观察结果、媒体文件等字段。此构建版本包含36个轨迹、906帧、6家公司,环境分布为browser-os(13个)、macos-web-next(14个)和windows-web-next(9个)。数据生成来源为SynthUX(synthux.multicompany.day.v1),观察结果是因果的(突变通道为低级别输入),在捕获过程中未使用高级别应用程序突变。
SynthUX Computer Usage Dataset (Economic Simulation r3) is a vision-based computer usage trajectory dataset generated by SynthUX. Each record represents a device session of a worker within a simulated company: a target is expanded into a node tree, where terminal nodes drive real desktop simulator applications such as terminals, Notes, VS Code, browsers, Slack/Teams, email clients, spreadsheets, slides, etc. via low-level mouse and keyboard inputs. The observed trajectories are recorded as screen videos and frames corresponding to each node. This dataset supports native multi-application execution per application (no workspace overlay), content diversity across participants, cross-OS messaging, and large-scale multi-company economic simulations. Each economic simulation is released as an independent dataset repository. The dataset includes data files (data.jsonl), screenshots of each terminal node (stored in the screenshots/ directory), and full screen recording videos (stored in the videos/ directory). The record schema includes fields such as trajectory ID, participant information, environment, target, node tree, input events, observations, media files, and more. This build contains 36 trajectories, 906 frames, and 6 companies, with environment distributions of browser-os (13 instances), macos-web-next (14 instances), and windows-web-next (9 instances). The data generation source is SynthUX (synthux.multicompany.day.v1), and the observations are causal (the mutation channel is low-level inputs), with no high-level application mutations used during the capture process.




