微软联合发布 MiniCorp,一个用于生成企业级智能体训练与评估数据的办公室模拟器:智能体团队运营电商公司,通过消息和会议决策定价、库存与广告投放,模拟市场中的顾客与自适应竞争对手会做出反应。该市场经真实市场研究校验,四周 20% 折扣使销量提升 71%,价格恢复后销量仍比对照组高 31%。检查点机制可让同一市场状态用不同决策重放,产出真实企业记录无法提供的配对结果。
New paper from Microsoft and colleagues on generating enterprise data for agents.
If you train or evaluate agents on long-running business work, it's worth your time.
(bookmark it)
MiniCorp is an office simulator. A firm of agents runs an e-commerce company, discusses decisions in messages and meetings, and sets prices, inventory, and ad spend. A simulated market with customers and adaptive competitors responds, and the firm reacts to the results.
The market is checked against published studies of real ones. A four-week 20% discount lifts units by 71%, and sales stay 31% above control after prices return to normal.
Checkpoints let the same market state replay with a different decision, giving paired outcomes that real company records cannot provide.
Paper: https://arxiv.org/abs/2610.05912
Chat with Paper: https://academy.dair.ai/papers/minicorp-the-last-mile-of-the-ai-agent-firm-2610.05912
来源:DAIR.AI · x.com