ResearchBenchmarkstask3 on puzzle 4x5Follow100Success RateTRL-4235077Oct 26, 2025Evaluation ResultsMethodMethodLinksSuccess RateTRL2025.10100MC2025.108TD2025.102FBC2025.100IQL2025.100CRL2025.100SAC+BC2025.100QRL2025.100TDP2025.100COE2025.100