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

TextCraft

Benchmarks

Task NameDataset NameSOTA ResultTrend
Interactive Decision MakingTextcraft
Success Rate99.6
60
Item craftingTextCraft (test)
Success Rate71
32
Language Agent TaskTextCraft
Success Rate (SR)100
17
One-step next-observation predictionTextCraft (test)
Token F195
16
Hierarchical CraftingTextCraft
Score88.9
9
Compositional planningTextCraft
Success Rate94
8
Autonomous ExplorationTextCraft
Steps8.7
7
Task ExecutionTEXTCRAFT-SYNTH 8K context Easy (evaluation)
Success Rate100
4
Sequential CraftingTextCraft-4
Success Rate (SR)45.5
4
Sequential CraftingTextCraft-3
Success Rate (%)82.5
4
Sequential CraftingTextCraft 2
Success Rate (SR)94.3
4
Crafting ace itemsTextCraft
Success Rate58
4
Showing 12 of 12 rows