arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2003.03124cs.LGcs.NEstat.ML
收藏

通过学习记忆寻找在线神经更新规则

Finding online neural update rules by learning to remember

DeepMind(深度思维)

Karol Gregor

摘要:

我们从头开始研究神经激活(身体)和权重(突触)的在线局部更新规则的学习。我们用小向量表示每个权重和激活的状态,并使用(Meta)神经网络参数化它们的更新。不同的神经元类型由不同的嵌入向量表示,这允许对所有神经元使用相同的两个函数。我们不是像在类似系统中通常做的那样,直接使用进化或长期反向传播来训练目标,而是激励和研究一个不同的目标:记住过去的经验片段。我们解释了这个目标与标准的反向传播训练和其他形式的学习的关系。我们使用短期反向传播为这个目标进行训练,并分析作为不同网络类型和问题难度的函数的性能。我们发现,这种分析提供了有趣的见解是什么构成了一个学习规则。我们还讨论了如何这样的系统可以形成一个自然的基板,解决主题,如情景记忆,元学习和辅助目标。

英文摘要:

We investigate learning of the online local update rules for neural activations (bodies) and weights (synapses) from scratch. We represent the states of each weight and activation by small vectors, and parameterize their updates using (meta-) neural networks. Different neuron types are represented by different embedding vectors which allows the same two functions to be used for all neurons. Instead of training directly for the objective using evolution or long term back-propagation, as is commonly done in similar systems, we motivate and study a different objective: That of remembering past snippets of experience. We explain how this objective relates to standard back-propagation training and other forms of learning. We train for this objective using short term back-propagation and analyze the performance as a function of both the different network types and the difficulty of the problem. We find that this analysis gives interesting insights onto what constitutes a learning rule. We also discuss how such system could form a natural substrate for addressing topics such as episodic memories, meta-learning and auxiliary objectives.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑