Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Game-based visual-agent task completion on VAB-Minecraft

73.28Success Rate

Gemini 3 Flash

21.2834.7848.2861.78May 13, 2026
Updated 20d ago

Evaluation Results

MethodLinks
2026.05
73.280.7884
2026.05
68.960.7541
2026.05
68.10.7495
2026.05
67.240.7462
2026.05
62.070.7114
2026.05
61.20.6938
2026.05
56.030.6701
2026.05
55.170.6634
2026.05
52.590.6308
2026.05
38.790.4668
2026.05
29.310.3754
2026.05
23.280.3017