微软与伊利诺伊大学开发 StudentSim 模拟学生帮助 AI 辅导员学习
Simulated students that make realistic mistakes help AI tutors learn faster
微软和大学搞了个叫 StudentSim 的东西，能模拟学生犯真实错误，让 AI 辅导员学得更快，还比 GPT-5.4 强。
微软和伊利诺伊大学开发的 StudentSim 能从少量数据中模拟出具有真实错误的学生，为 AI 辅导员提供快速低成本反馈。在涵盖国际象棋、英语和数学的测试中，它表现优于 GPT-5.4。用 StudentSim 训练的国际象棋 AI 辅导员在三个版本中获得了最高专家评分。
Simulated students that make realistic mistakes help AI tutors learn faster
Microsoft and the University of Illinois built StudentSim to replicate individual students from limited data and give AI tutors fast, low-cost feedback. In tests covering 60 students across chess, English, and math, it outperformed GPT-5.4. A chess tutor trained with StudentSim also earned the highest expert ratings among three versions tested. The article Simulated students that make realistic mistakes help AI tutors learn faster appeared first on The Decoder .
