Reproduce “Q-Shaped Options for Hierarchical Reinforcement Learning” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Clarisse Wibault, Antoine Gorceix, Antonio Léon Villares, Alexey Zakharov, Evangelos Chatzaroulas, Michael Matthews, Eduardo Pignatelli, Jakob Foerster
QSO learns hierarchical options for offline, long-horizon goal-reaching. Each option encodes a segment’s starting and achieved states; shared training with low-level and high-level Q functions preserves control-relevant distinctions while encouraging aggregation. Across 11 OGBench environments, QSO averaged 54% success rate, versus 49% for DSHARSA, after 1M training steps. Results cover offline datasets and simulated environments.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.