ResearchBenchmarkstask5 on puzzle 4x5Follow8,800Success RateTRL-3522,0244,4006,776Oct 26, 2025Evaluation ResultsMethodMethodLinksSuccess RateTRL2025.108,800MC2025.10800TD2025.10200FBC2025.100IQL2025.100CRL2025.100SAC+BC2025.100QRL2025.100TDP2025.100COE2025.100