Organizing recurrent network dynamics by task-computation to enable continual learning
Organizing recurrent network dynamics by task-computation to enable continual learning
复制标题
DOI：
--
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Lea Duncker;Laura N. Driscoll;K. Shenoy;M. Sahani;David Sussillo
中科院分区：
文献类型：
--
作者：
Lea Duncker;Laura N. Driscoll;K. Shenoy;M. Sahani;David Sussillo
生物系统面临着需要不断学习的动态环境。这些系统如何平衡学习的灵活性和对先前行为的记忆的鲁棒性之间的紧张关系，目前还不清楚。没有灾难性干扰的持续学习仍然是机器学习中的一个挑战性问题。在这里，我们开发了一种新的学习规则，旨在最大限度地减少递归网络中顺序学习任务之间的干扰。我们的学习规则在用于先前学习任务的活动定义的子空间内保留了网络动态。它鼓励与新任务相关的动态，否则可能会干扰探索正交子空间，并且它允许在可能的情况下重用先前建立的动态基序。采用神经科学中使用的一组任务，我们证明了我们的方法成功地消除了灾难性的干扰，并提供了一个显着的改进，比以前的持续学习算法。使用动态系统分析，我们证明了使用我们的方法训练的网络可以在类似的任务中重用类似的动态结构。这种共享计算的可能性允许在顺序训练期间更快地学习。最后，我们确定了组织的差异时出现的培训任务顺序与同时进行。
Biological systems face dynamic environments that require continual learning. It is not well understood how these systems balance the tension between flexibility for learning and robustness for memory of previous behaviors. Continual learning without catastrophic interference also remains a challenging problem in machine learning. Here, we develop a novel learning rule designed to minimize interference between sequentially learned tasks in recurrent networks. Our learning rule preserves network dynamics within activity-defined subspaces used for previously learned tasks. It encourages dynamics associated with new tasks that might otherwise interfere to instead explore orthogonal subspaces, and it allows for reuse of previously established dynamical motifs where possible. Employing a set of tasks used in neuroscience, we demonstrate that our approach successfully eliminates catastrophic interference and offers a substantial improvement over previous continual learning algorithms. Using dynamical systems analysis, we show that networks trained using our approach can reuse similar dynamical structures across similar tasks. This possibility for shared computation allows for faster learning during sequential training. Finally, we identify organizational differences that emerge when training tasks sequentially versus simultaneously.