X-IL: Exploring the Design Space of Imitation Learning Policies
This work addresses the problem of navigating complex design choices in imitation learning for practitioners and researchers, though it is incremental as it builds on existing methods through systematic exploration.
The authors tackled the challenge of exploring the vast design space in imitation learning by introducing X-IL, an open-source framework that enables systematic experimentation with policy components, leading to novel configurations that outperform existing methods on robot learning benchmarks with significant performance gains.
Designing modern imitation learning (IL) policies requires making numerous decisions, including the selection of feature encoding, architecture, policy representation, and more. As the field rapidly advances, the range of available options continues to grow, creating a vast and largely unexplored design space for IL policies. In this work, we present X-IL, an accessible open-source framework designed to systematically explore this design space. The framework's modular design enables seamless swapping of policy components, such as backbones (e.g., Transformer, Mamba, xLSTM) and policy optimization techniques (e.g., Score-matching, Flow-matching). This flexibility facilitates comprehensive experimentation and has led to the discovery of novel policy configurations that outperform existing methods on recent robot learning benchmarks. Our experiments demonstrate not only significant performance gains but also provide valuable insights into the strengths and weaknesses of various design choices. This study serves as both a practical reference for practitioners and a foundation for guiding future research in imitation learning.