Search papers, labs, and topics across Lattice.
1
0
2
Public ARC-AGI-3 games, intended to test intelligent exploration, are surprisingly solvable with trivial heuristics, rendering them a poor benchmark.
