Learning to Rehearse in Long Sequence Memorization
This addresses efficient reasoning on long sequences with limited storage for applications such as question answering and recommendation, but it appears incremental as it builds on existing memory augmented networks.
The paper tackles the problem of forgetting early information and treating all contents equally in memory augmented neural networks for long-sequence memorization, proposing a Rehearsal Memory with self-supervised rehearsal and a history sampler that achieves improved performance on tasks like bAbI and text/video question answering.
Existing reasoning tasks often have an important assumption that the input contents can be always accessed while reasoning, requiring unlimited storage resources and suffering from severe time delay on long sequences. To achieve efficient reasoning on long sequences with limited storage resources, memory augmented neural networks introduce a human-like write-read memory to compress and memorize the long input sequence in one pass, trying to answer subsequent queries only based on the memory. But they have two serious drawbacks: 1) they continually update the memory from current information and inevitably forget the early contents; 2) they do not distinguish what information is important and treat all contents equally. In this paper, we propose the Rehearsal Memory (RM) to enhance long-sequence memorization by self-supervised rehearsal with a history sampler. To alleviate the gradual forgetting of early information, we design self-supervised rehearsal training with recollection and familiarity tasks. Further, we design a history sampler to select informative fragments for rehearsal training, making the memory focus on the crucial information. We evaluate the performance of our rehearsal memory by the synthetic bAbI task and several downstream tasks, including text/video question answering and recommendation on long sequences.