Neural networks optimized to perform computational tasks, which were proposed as hypotheses of the proprioceptive system, develop representations that generalize from synthetic data to predict neural dynamics in CN and S1 of primates.

中文翻译：

---

任务驱动的神经网络模型预测本体感觉的神经动力学

为执行计算任务而优化的神经网络（作为本体感觉系统的假设而提出）开发了从合成数据中概括的表示，以预测灵长类动物 CN 和 S1 的神经动力学。