Code and Data for the KR 2023 paper "Learning General Policies with Policy Gradient Methods"

Name: Code and Data for the KR 2023 paper "Learning General Policies with Policy Gradient Methods"
Keywords: classical planning, automated planning, deep learning, graph neural networks, generalized planning, general policies

Ståhlberg, Simon; Bonet, Blai; Geffner, Hector

Found an issue? Give us feedback

ZENODOarrow_drop_down

ZENODO

Software . 2023

License: https://www.gnu.org/licenses/agpl.txt

Data sources: Datacite

ZENODO

Software . 2023

License: https://www.gnu.org/licenses/agpl.txt

Data sources: Datacite

Code and Data for the KR 2023 paper "Learning General Policies with Policy Gradient Methods"

integration_instructionsResearch softwarekeyboard_double_arrow_right Software 01 Jun 2023 English Publisher:Zenodo

Authors: Ståhlberg, Simon; Bonet, Blai; Geffner, Hector;

doi: 10.5281/zenodo.7993858 , 10.5281/zenodo.7993859

Code and Data for the KR 2023 paper "Learning General Policies with Policy Gradient Methods"

- Summary
- Subjects
- Metrics

Abstract

This archive contains three files: The file 'Code.zip' contains the source code we used to train and test models. Please refer to the included README.md file for additional information. The file 'Domains.zip' contains the PDDL files of the domains that we used in the paper. The test instances can be found in the subdirectory 'test' for each domain. The file 'Models.zip' contains two models for each domain: the one that had the best performance on the validation set, and the latest one. In addition to the models, there are training and test logs. A test log is the output of the planner using the model to solve an instance.

Related Organizations

Keywords

classical planning, automated planning, deep learning, graph neural networks, generalized planning, general policies

Impact byBIP!

	selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).	0
	popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.	Average
	influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).	Average
	impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.	Average