Code and Data for AIJ submission

Code and data for the article submitted to the Artificial Intelligence journal (AIJ), titled "Automated Planning Instance Generation with Neuro-Symbolic AI", which extends our previous ECAI paper with additional experimentation.

Experiment results can be found in experiments (for NeSIG and ablations) and in data/instance_generators for ad-hoc generators.
Each subfolder experiments/<experiment_id> contains a file experiment_info.json with the parameters used to run the experiment.
NeSIG logs and checkpoints are not provided due to their large size.

Code and data for ECAI24

Code and data for the ECAI24 paper Nesig: A neuro-symbolic method for learning to generate planning problems.
Experiment results can be found in experiments (for NeSIG and ablations) and in data/instance_generators for ad-hoc generators.
Each subfolder experiments/<experiment_id> contains a file experiment_info.json with the parameters used to run the experiment.
NeSIG logs and checkpoints are not provided due to their large size.