Paper

MaxIM: Maximally Informative Incremental Summarization via Reinforcement Learning · ByteTrail