List your company
ResearchFormerly WizWand

Interactive Instruction Following on ALFWorld Unseen

86.68Success RateCOMAP
-2.687220.513943.71566.9161Jun 1, 2026Jun 3, 2026Jun 6, 2026Jun 9, 2026Jun 11, 2026Jun 14, 2026Jun 17, 2026
Updated 4mo ago

Evaluation results

MethodLinks
COMAPBackbone=Qwen3-8B, Category=Co-evolving, Training-based=true2026.0686.68—
Imagine-and-ActBackbone=GPT-5.4, Category=API-Called Models, Training-free=true2026.0684.7—
Imagine-and-ActBackbone=DeepSeek-V4-Pro, Category=API-Called Models, Training-free=true2026.0683.9—
ITPRBackbone=Qwen3-8B, Category=World Modeling, Training-free=true2026.0683.2—
ReActBackbone=GPT-5.4, Category=API-Called Models, Training-free=true2026.0680.6—
ReActBackbone=DeepSeek-V4-Pro, Category=API-Called Models, Training-free=true2026.0680.2—
IWMBackbone=Qwen3-8B, Category=World Modeling, Training-free=true2026.0679.5—
Entropy-Aware OPDStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.0679.116.75
WKMBackbone=Qwen3-8B, Category=World Modeling, Training-free=true2026.0676.87—
OPDStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.0674.8517.16
SAGE-OPDStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.0673.8817.57
TCOD-F2BStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.0673.1317.87
OPD (rollout 2)Student Model=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.0672.3918.37
Imagine-and-Act (w/ SFT)Backbone=Qwen3-8B, Category=Base Agent Planning, Training-based=true2026.0672.34—
SAGE-OPDStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-32B2026.0670.917.98
COMAPBackbone=Qwen3-4B, Category=Co-evolving, Training-based=true2026.0670.68—
OPD (rollout 2)Student Model=Qwen3-0.6B, Teacher Model=Qwen3-32B2026.0670.3718.21
SAGE-OPDStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-8B2026.0670.1517.86
BaseTeacher Model=Qwen3-32B2026.0669.418.21
OPDStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-32B2026.0668.6618.65
Imagine-and-ActBackbone=DeepSeek-V4-Flash, Category=API-Called Models, Training-free=true2026.0668.3—
Entropy-Aware OPDStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-32B2026.0668.1318.28
TCOD-F2BStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-32B2026.0667.9119.26
BaseTeacher Model=Qwen3-8B2026.0667.1617.24
Entropy-Aware OPDStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-8B2026.0665.6718.2
OPD (rollout 2)Student Model=Qwen3-1.7B, Teacher Model=Qwen3-8B2026.0664.9317.87
ReActBackbone=DeepSeek-V4-Flash, Category=API-Called Models, Training-free=true2026.0664.5—
SAGE-OPDStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-8B2026.0661.9418.89
OPDStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-8B2026.0661.9417.67
TCOD-F2BStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-8B2026.0661.1918.82
OPD (rollout 2)Student Model=Qwen3-0.6B, Teacher Model=Qwen3-8B2026.0659.720.31
TCOD-F2BStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-8B2026.0658.2119.16
ReAct (w/ SFT)Backbone=Qwen3-8B, Category=Base Agent Planning, Training-based=true2026.0657.62—
OPDStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-8B2026.0656.7220.3
Entropy-Aware OPDStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-8B2026.0654.4820.75
ITPRBackbone=Qwen3-4B, Category=World Modeling, Training-free=true2026.0648.54—
Imagine-and-Act (w/ SFT)Backbone=Qwen3-4B, Category=Base Agent Planning, Training-based=true2026.0648.4—
IWMBackbone=Qwen3-4B, Category=World Modeling, Training-free=true2026.0647.2—
ReAct (w/ SFT)Backbone=Qwen3-4B, Category=Base Agent Planning, Training-based=true2026.0645.8—
WKMBackbone=Qwen3-4B, Category=World Modeling, Training-free=true2026.0644.9—
ITPIBackbone=Qwen3-8B, Category=Test-time Evolving, Training-free=true2026.0639.2—
Imagine-and-ActBackbone=Qwen3-8B, Category=Base Agent Planning, Training-free=true2026.0630.55—
ITPIBackbone=Qwen3-4B, Category=Test-time Evolving, Training-free=true2026.0628.7—
RAPBackbone=Qwen3-8B, Category=Test-time Evolving, Training-free=true2026.0626.8—
BaseStudent Model=Qwen3-1.7B2026.0626.1224.72
SFTStudent Model=Qwen3-1.7B, Teacher Model=Qwen3-8B2026.0626.1224.54
ReflexionBackbone=Qwen3-8B, Category=Test-time Evolving, Training-free=true2026.0622.4—
RAPBackbone=Qwen3-4B, Category=Test-time Evolving, Training-free=true2026.0622.1—
ReflexionBackbone=Qwen3-4B, Category=Test-time Evolving, Training-free=true2026.0619.2—
Imagine-and-ActBackbone=Qwen3-4B, Category=Base Agent Planning, Training-free=true2026.0618.2—
ReActBackbone=Qwen3-8B, Category=Base Agent Planning, Training-free=true2026.0617.85—
ReActBackbone=Qwen3-4B, Category=Base Agent Planning, Training-free=true2026.0612.5—
SFTStudent Model=Qwen3-0.6B, Teacher Model=Qwen3-8B2026.064.4828.87
BaseStudent Model=Qwen3-0.6B2026.060.7529.82