List your company

Decision Making on ALFWorld, WebShop, ScienceWorld, and BabyAI (test)

66.42ALFWorld Success RateNPM + Workflows
22.178433.664245.1556.6358Jun 29, 2026

Evaluation results

MethodLinks
NPM + WorkflowsModel=Qwen3-8B, Paradigm=Hybrid2026.0666.4253.9431.8915.3141.89
WorkflowsModel=Qwen3-8B, Paradigm=Explicit2026.0662.6948.9329.5710.437.9
NPM + WorkflowsModel=Qwen3-4B, Paradigm=Hybrid2026.0661.1947.8421.9719.3837.6
NPMModel=Qwen3-8B, Paradigm=Implicit2026.0656.7248.2426.0714.2636.32
WorkflowsModel=Qwen3-4B, Paradigm=Explicit2026.0652.9945.7323.5914.634.23
Mass-MeanModel=Qwen3-8B, Paradigm=Implicit2026.0649.2544.2625.328.7131.89
CAAModel=Qwen3-8B, Paradigm=Implicit2026.0647.0120.669.1714.3122.79
InsightsModel=Qwen3-8B, Paradigm=Explicit2026.0646.2745.2228.097.7731.84
InsightsModel=Qwen3-4B, Paradigm=Explicit2026.0644.0340.7222.5912.0929.86
NPMModel=Qwen3-4B, Paradigm=Implicit2026.0640.34818.219.0431.39
No MemoryModel=Qwen3-8B, Paradigm=None2026.0639.5546.2524.9811.7430.63
NPM + WorkflowsModel=MiniCPM3-4B, Paradigm=Hybrid2026.0638.8151.2110.9736.934.47
WorkflowsModel=MiniCPM3-4B, Paradigm=Explicit2026.0632.8454.9710.532.3932.68
NPMModel=MiniCPM3-4B, Paradigm=Implicit2026.0631.3443.538.731.9128.87
No MemoryModel=Qwen3-4B, Paradigm=None2026.0630.644.8118.418.7528.14
Mass-MeanModel=Qwen3-4B, Paradigm=Implicit2026.0629.139.0718.1617.8226.04
InsightsModel=MiniCPM3-4B, Paradigm=Explicit2026.0628.3647.961027.7828.53
Mass-MeanModel=MiniCPM3-4B, Paradigm=Implicit2026.0626.8734.218.0229.7324.71
CAAModel=Qwen3-4B, Paradigm=Implicit2026.0626.8725.5211.483.4816.84
No MemoryModel=MiniCPM3-4B, Paradigm=None2026.0623.8832.217.292722.6
CAAModel=MiniCPM3-4B, Paradigm=Implicit2026.0623.8813.856.2930.2918.58