Reproduce “AI Agents for the Dhumbal Card Game: A Comparative Study” 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
Sahaj Raj Malla
The authors compare hand-coded, search-based, learned, and random AI strategies for Dhumbal, a hidden-information card game. The Aggressive agent uses rules for when to call Jhyap and which cards to discard or take. In a four-player, 1,024-round simulation, it won 88.3% of rounds; this result covers selected agents and does not establish performance against people or other game settings.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.