English

Multitask Soft Option Learning

Machine Learning 2020-06-23 v3 Machine Learning

Abstract

We present Multitask Soft Option Learning(MSOL), a hierarchical multitask framework based on Planning as Inference. MSOL extends the concept of options, using separate variational posteriors for each task, regularized by a shared prior. This ''soft'' version of options avoids several instabilities during training in a multitask setting, and provides a natural way to learn both intra-option policies and their terminations. Furthermore, it allows fine-tuning of options for new tasks without forgetting their learned policies, leading to faster training without reducing the expressiveness of the hierarchical policy. We demonstrate empirically that MSOL significantly outperforms both hierarchical and flat transfer-learning baselines.

Keywords

Cite

@article{arxiv.1904.01033,
  title  = {Multitask Soft Option Learning},
  author = {Maximilian Igl and Andrew Gambardella and Jinke He and Nantas Nardelli and N. Siddharth and Wendelin Böhmer and Shimon Whiteson},
  journal= {arXiv preprint arXiv:1904.01033},
  year   = {2020}
}

Comments

Published at UAI 2020

R2 v1 2026-06-23T08:25:54.230Z