Reproduce “Dominion: A New Frontier for AI Research” 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
Danny Halawi, Aron Sarmasi, Siena Saltzen, Joshua McCoy
The authors present Dominion as a benchmark for adapting card-buying strategies to changing game setups and evaluate a Rainbow-DQN buyer in two-player, second-edition base-game matches. The buyer selects cards to gain, while Provincial heuristics handle action-card decisions. Against Provincial, it records 575 wins, 62 ties, and 363 losses in 1,000 games; results do not establish transfer across setups or human-level play.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.