Augmenting Language Models with Long-Term Memory Weizhi Wang, Li Dong

To address the length limit issue, the most straightforward method is to simply scale up the input context length.