ResearchBenchmarkstask2 on puzzle 4x5Follow9,900Success RateTRL-3962,2774,9507,623Oct 26, 2025Evaluation ResultsMethodMethodLinksSuccess RateTRL2025.109,900MC2025.106,700TD2025.10400FBC2025.100IQL2025.100CRL2025.100SAC+BC2025.100QRL2025.100TDP2025.100COE2025.100