XOR game
Benchmarks
TaskDatasetSOTA resultTrendResults
XOR game Generalization: Train n=2, k=3; Eval n=3, k=3
0Success Rate (Greedy Policy)
10
XOR game n=k=2 (train eval)
100Success Rate (Greedy Policy pi*)
10
XOR game n=k=3 (train eval)
—Primary metric
0