Train one model — it defends the target, and attacks it. Then climb the ladder.
heuristicOne vs heuristicOne — the model to beat. Open viewer ›
Full ladder ›