List your company

Interactive Task Solving on ALFWorld

63.65Average task-level performanceATG
0.875617.172833.4749.7672Jul 2, 2026

Evaluation results

MethodLinks
ATGBackbone=Llama-3-8B2026.0763.65
ATGBackbone=Gemma-7B2026.0758.71
ATGBackbone=Mistral-7B2026.0755.73
ReActBackbone=GPT-42026.0741.24
PoGBackbone=Gemma-7B2026.0732.85
ToTBackbone=Gemma-7B2026.0725.91
PoGBackbone=Mistral-7B2026.0723.72
PoGBackbone=Llama-3-8B2026.0721.17
ToTBackbone=Mistral-7B2026.0718.25
ToTBackbone=Llama-3-8B2026.0715.33
CAMELBackbone=Mistral-7B2026.0711.68
ReflexionBackbone=Mistral-7B2026.078.84
CAMELBackbone=Llama-3-8B2026.078.76
CAMELBackbone=Gemma-7B2026.077.66
ReActBackbone=GPT-3.5-Turbo2026.077.3
ReActBackbone=Mistral-7B2026.076.57
ReflexionBackbone=Gemma-7B2026.075.11
ReActBackbone=Gemma-7B2026.074.38
ReflexionBackbone=Llama-3-8B2026.074.38
ReActBackbone=Llama-3-8B2026.073.29