Reusable Test Suites for Reinforcement Learning

Reinforcement learning (RL) agents show great promise in solving sequential decision-making tasks. However, validating the reliability and performance of the agent policies'behavior for deployment remains challenging. Most reinforcement learning policy testing methods produce test suites tailored to the agent policy being tested, and their relevance to other policies is unclear. This work presents Multi-Policy Test Case Selection (MPTCS), a novel automated test suite selection method for RL environments, designed to extract test cases generated by any policy testing framework based on their solvability, diversity, and general difficulty. MPTCS uses a set of policies to select a diverse collection of reusable policy-agnostic test cases that reveal typical flaws in the agents'behavior. The set of policies selects test cases from a candidate pool, which can be generated by any policy testing method, based on a difficulty score. We assess the effectiveness of the difficulty score and how the method's effectiveness and cost depend on the number of policies in the set. Additionally, a method for promoting diversity in the test suite, a discretized general test case descriptor surface inspired by quality-diversity algorithms, is examined to determine how it covers the state space and which policies it triggers to produce faulty behaviors.

• Publication year

  2025

• Venue

  International Conference on Testing Software and Systems

• Publication date

  2025-08-29

• Fields of study

  Computer Science

• Identifiers
  DOI 10.48550/arXiv.2508.21553arXiv 2508.21553
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Diversity-Oriented Testing for Competitive Game Agent via Constraint-Guided Adversarial Agent Training

  2025cited by this paper
• Provably Robust Federated Reinforcement Learning

  2025cited by this paper
• Book review: Christoph Molnar. 2020. Interpretable Machine Learning: A Guide for Making Black Box Models Explainable

  2024cited by this paper
• Testing for Fault Diversity in Reinforcement Learning

  2024influential reference
• Test Where Decisions Matter: Importance-driven Testing for Deep Reinforcement Learning

  2024cited by this paper
• SMARLA: A Safety Monitoring Approach for Deep Reinforcement Learning Agents

  2023cited by this paper
• Pgx: Hardware-Accelerated Parallel Game Simulators for Reinforcement Learning

  2023cited by this paper
• Generative Model-Based Testing on Decision-Making Policies

  2023cited by this paper
• Testing of Deep Reinforcement Learning Agents with Surrogate Models

  2023cited by this paper
• Automatic Metamorphic Test Oracles for Action-Policy Testing

  2023cited by this paper
• Semantic‐aware two‐phase test case prioritization for continuous integration

  2023cited by this paper
• Explainable Reinforcement Learning: A Survey and Comparative Review

  2023cited by this paper
• A Search-Based Testing Approach for Deep Reinforcement Learning Agents

  2022cited by this paper
• Explainable Deep Reinforcement Learning: State of the Art and Challenges

  2022cited by this paper
• Search-Based Testing of Reinforcement Learning

  2022cited by this paper
• COOL-MC: A Comprehensive Tool for Reinforcement Learning and Model Checking

  2022cited by this paper
• Many-Objective Reinforcement Learning for Online Testing of DNN-Enabled Systems

  2022cited by this paper
• Single and Multi-objective Test Cases Prioritization for Self-driving Cars in Virtual Environments

  2021cited by this paper
• MDPFuzz: testing models solving Markov decision processes

  2021influential reference
• MinAtar: An Atari-Inspired Testbed for Thorough and Reproducible Reinforcement Learning Experiments

  2019influential reference
• Reinforcement learning for automatic test case prioritization and selection in continuous integration

  2017cited by this paper
• Safe Reinforcement Learning via Shielding

  2017cited by this paper
• Proximal Policy Optimization Algorithms

  2017cited by this paper
• Empirical evaluation of mutation‐based test case prioritization techniques

  2017cited by this paper
• Mastering the game of Go with deep neural networks and tree search

  2016cited by this paper
• Illuminating search spaces by mapping elites

  2015influential reference
• Comparing and combining test-suite reduction and regression test selection

  2015cited by this paper
• A fast and elitist multiobjective genetic algorithm: NSGA-II

  2002cited by this paper
• Reinforcement Learning: An Introduction

  1998cited by this paper

• No citing papers are available for this paper.