ResearchBenchmarksReinforcement Learning on POPGym Aggregated (48 tasks)Reinforcement Learning on POPGym Aggregated (48 tasks)Follow10.4Aggregated Return (All)ELMUR-13.104-7.002-0.95.202Oct 8, 2025Evaluation resultsMethodMethodLinksAggregated Return (All)Aggregated Return (Puzzle)Aggregated Return (Reactive)ELMUR2025.1010.41.29.2RATE2025.109.50.459.1BC-LSTM2025.109-0.29.1DT2025.105.8-3.59.3BC-MLP2025.10-6.8-11.95.1Rand.2025.10-12.2-14.62.3