Reference library
RL2: Fast Reinforcement Learning via Slow Reinforcement Learning
Paper summary
RL-squared trains a recurrent network through a slower reinforcement-learning procedure so its runtime hidden state implements rapid adaptation. Observations, actions, rewards, and episode indicators are inputs to the learned algorithm.
Relevance to the project
A comparison for storing adaptive behavior in state rather than only weight changes. Outer training, task information, and inner adaptation must be recorded separately.
The source supplies rewards and a task distribution.
Source record
- Reference ID
- arxiv-1611-02779
- Archived version
- arXiv preprint; retrieved version indicated in PDF
- Summary provenance
- reference-library/papers/arxiv-1611-02779/metadata.json · summary
- Review status
- initial orientation summary; not a critical full-paper review
- Reading scope
- Selected abstract/opening and available closing discussion/conclusion passages in the preserved text extract; broader project notes provide context. No reproduction of experiments, exhaustive methods review, or equation/figure audit.