Skip to content

Code for my paper: Meta-Learning of Compositional Task Distributions in Humans and Machines

Notifications You must be signed in to change notification settings

sreejank/Compositional_MetaRL

Repository files navigation

Compositional_MetaRL

Code for Kumar et al. 2020 "Meta-Learning of Compositional Task Distributions in Humans and Machines" (https://arxiv.org/abs/2010.02317)

Contents:

  • grid_grammar.py: Implements the generative grammar and provides functions to generate boards from compositional and null task distributions.

  • grid_env.py: Reinforcement learning enviornment (OpenAI Gym) that implements task in the paper.

  • null_task_distribution.py: Code to train fully connected network to learn conditional distributions within the compositional boards and perform Gibbs sampling to obtain null task distribution.

  • train.py: Code to train the meta-reinforcement learning agent on the grid task described in the paper.

  • held_out/: Directory containing example boards in each distribution

    • all.npy: Compositional boards

    • null.npy: Null boards

About

Code for my paper: Meta-Learning of Compositional Task Distributions in Humans and Machines

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages