Meta 发布 Muse Glimmer 技术报告:在模拟数字生活中训练智能体
Meta 把智能体放进模拟的通讯录、日历和消息堆里练多天任务,还把幻觉轨迹全扔掉,这篇报告写法挺实在。
Meta 的 Muse Glimmer 技术报告介绍了在仿真数字环境中训练智能体的做法。每个模拟场景包含 10–14 个联系人、数周的聊天记录、日历和文档,智能体在这些环境中练习跨越多天的任务。训练时会把含幻觉结果或未经核实成功声称的轨迹直接丢弃,以此控制数据质量。
Metas Muse Glimmer tech report is pretty interesting.
Metas Muse Glimmer report describes training agents inside simulated digital lives, complete with 10–14 contacts, weeks of messages, calendars and documents. The agents practice tasks spanning multiple days, while training trajectories containing hallucinated results or unverified claims of success are discarded.