R2-Dreamer: Redundancy-Reduced World Models without Decoders or Augmentation

R2-Dreamer: Redundancy-Reduced World Models without Decoders or Augmentation explores A decoder-free model-based reinforcement learning framework that uses an internal redundancy-reduction objective to learn efficient representations without data augmentation, achieving competitive performance and faster training.. Commercial viability score: 7/10 in Reinforcement Learning.