arXiv · 2611.17793
MASTARS: Multi-Agent Sequential Trajectory Augmentation with Return-Conditioned Subgoals
Abstract
The performance of offline Multi-Agent Reinforcement Learning (MARL) is limited by the quality of the fixed offline dataset, motivating trajectory augmentation. However, augmenting multi-agent trajectories presents a fundamental challenge of the diversity-coordination trade-off: independent generation improves diversity but lacks coordination, while joint generation induces coordination but lacks diversity, reproducing the training distribution. To address this challenge in multi-agent data augmentation, we introduce MASTARS, a novel diffusion-based framework that generates multi-agent trajectories sequentially across agents while enforcing cross-agent consistent coordination via the method of inpainting. MASTARS produces trajectories that are diverse, coordinated and globally coherent. Experiments on various benchmarks such as MPE, SMAC, and SMAC-v2 demonstrate significant performance gains across multiple offline MARL methods.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Jiwon Jeon, Myungsik Cho, Woojun Kim, Seongmin Kim, Woohyeon Byeon, Seungyul Han, Youngchul Sung. 2026-11-13. MASTARS: Multi-Agent Sequential Trajectory Augmentation with Return-Conditioned Subgoals. https://searcharxiv.com/abs/2611.17793
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Searcharxiv by Indiciae: BibTeX · RIS.