# Why are GRU and LSTM better than standard RNNs?

It seems that older RNNs have a limitation for their use cases and have been outperformed by other recurrent architectures, such as the LSTM and GRU.