A Large Recurrent Action Model: xLSTM enables Fast Inference for Robotics Tasks
ELLIS Unit / University Linz · LIT AI Lab / JKU Linz · Amazon · Meta FAIR & ELLIS Unit / JKU Linz · ELLIS Unit Linz / Johannes Kepler University Linz / NXAI GmbH · Johannes Kepler University (JKU) Linz · Johannes Kepler University Linz Austria · (Google) DeepMind · ELLIS Unit Linz, LIT AI Lab, Institute for Machine Learning, Johannes Kepler University, Institute for Advanced Research in Artificial Intelligence (IARAI)
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
In recent years, there has been a trend in the field of Reinforcement Learning (RL) towards large action models trained offline on large-scale datasets via sequence modeling. Existing models are primarily based on the Transformer architecture, which results in powerful agents. However, due to slow inference times, Transformer-based approaches are impractical for real-time applications, such as robotics. Recently, modern recurrent architectures, such as xLSTM and Mamba, have been proposed that exhibit parallelization benefits during training similar to the Transformer architecture while offering fast inference. In this work, we study the aptitude of these modern recurrent architectures for large action models. Consequently, we propose a Large Recurrent Action Model (LRAM) with an xLSTM at its core that comes with linear-time inference complexity and natural sequence length extrapolation abilities. Experiments on 432 tasks from 6 domains show that LRAM compares favorably to Transformers in terms of performance and speed.