In-context Reinforcement Learning with Algorithm Distillation.
Michael LaskinLuyu WangJunhyuk OhEmilio ParisottoStephen SpencerRichie SteigerwaldDJ StrouseSteven Stenberg HansenAngelos FilosEthan A. BrooksMaxime GazeauHimanshu SahniSatinder SinghVolodymyr MnihPublished in: ICLR (2023)
Keyphrases
- reinforcement learning
- learning algorithm
- improved algorithm
- times faster
- theoretical analysis
- computational complexity
- optimal solution
- cost function
- high accuracy
- preprocessing
- detection algorithm
- optimization algorithm
- expectation maximization
- experimental evaluation
- dynamic programming
- worst case
- computationally efficient
- multi agent
- segmentation algorithm
- objective function
- probabilistic model
- particle swarm optimization
- k means
- search space
- tree structure
- monte carlo
- function approximation
- neural network