Work

On Picking Good Policies: Leveraging Action-Policy Testing in Policy Training

Year: 2025

Type: conference-paper

Abstract: Testing is a natural approach to assess the quality of learned action policies π. Prior work introduced policy testing in AI planning as searching for bugs in π, that is, states where π is sub-optimal with respect to a given testing objective. Beyond quality assurance, an obvious application of these methods is policy selection: given several π to choose from, we can use testing to select the "least buggy" one. Here, we integrate testing-based policy selection into the training process. This includes making more informed decisions when selecting the final policy after training, as well as choosing more promising intermediate policies during the training process. Our experiments with ASNets action policies show that integrating testing allows us to more reliably obtain good-quality policies.

Source: Proceedings of the International Conference on Automated Planning and Scheduling

DOI: 10.1609/icaps.v35i1.36116

Authors: Jan Eisenhut, Daniel Fišer, Isabel Valera, Jörg Hoffmann

Institutions: Saarland University, Aalborg University, Max Planck Institute for Software Systems, German Research Centre for Artificial Intelligence

Language: English

Topic: AI-based Problem Solving and Planning

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: diamond

Funders: European Commission, Deutsche Forschungsgemeinschaft, HORIZON EUROPE Framework Programme

Awards: G256000070, G3032849772, Trustworthy Planning and Scheduling with Learning and Explanations, TRR 248: Grundlagen verständlicher SoftwareSysteme für eine nachvollziehbare cyberphysische Welt