This repository implements the paper "Utilizing Skipped Frames in Action Repeats for Improving Sample Efficiency in Reinforcement Learning".

All of the dependencies are in the conda_env.yml file. They can be installed manually or with the following command:

conda env create -f conda_env.yml

Quick start: bash scripts/cartpole/run.sh

This code is implemented on top of SAC Pytorch.

This work was supported in part by the Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIT) (No. 2021-0-01381, Development of Causal AI through Video Understanding) and in part by the Institute of Information & communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (2022-0-00951, Development of Uncertainty-Aware Agents Learning by Asking Questions).

@ARTICLE{9793636,
  author={Luu, Tung M. and Nguyen, Thanh and Vu, Thang and Yoo, Chang D.},
  journal={IEEE Access}, 
  title={Utilizing Skipped Frames in Action Repeats for Improving Sample Efficiency in Reinforcement Learning}, 
  year={2022},
  volume={10},
  number={},
  pages={64965-64975},
  doi={10.1109/ACCESS.2022.3182107}
  }