Towards Better Laplacian Representation in Reinforcement Learning with Generalized Graph Drawing

Authors: Kaixin Wang, Kuangqi Zhou, Qixin Zhang, Jie Shao, Bryan Hooi, Jiashi Feng

ICML 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We validate this via comprehensive experiments on a set of gridworld and continuous control environments.
Researcher Affiliation Collaboration 1National University of Singapore 2City University of Hong Kong 3Byte Dance AI lab.
Pseudocode No The paper does not contain any clearly labeled pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an explicit statement or link indicating that the code for their methodology is open-source or publicly available.
Open Datasets Yes The gridworld environments are built with Mini Grid (Chevalier-Boisvert et al., 2018) and the continuous control environments are created with Py Bullet (Coumans & Bai, 2016 2019).
Dataset Splits No The paper mentions collecting trajectories for training and discusses evaluation, but it does not specify explicit train/validation/test dataset splits (e.g., percentages or counts) within the main text.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., CPU, GPU models, memory) used for running the experiments.
Software Dependencies No The paper mentions environments built with Mini Grid and Py Bullet, and the use of Deep Q-learning, but it does not list specific version numbers for software dependencies such as libraries, frameworks, or programming languages.
Experiment Setup No The paper mentions that 'More details about training setup can be found in the Appendix' and provides some high-level descriptions of the experimental approach (e.g., 'd = 10 for the dimension of the Laplacian representation', 'learned with Deep Q-learning'), but it does not specify a comprehensive list of hyperparameters or detailed system-level training settings in the main text.