List your company

Interactive Agent Control on ALFWorld

99.25Success RateSKILL-DISCO + CodeAct
44.920459.025273.1387.2348Jun 25, 2026

Evaluation results

MethodLinks
SKILL-DISCO + CodeActBackbone=GPT-4o2026.0699.253.228,4683680.0107
CodeActBackbone=GPT-4o2026.0696.273.637,7215620.0109
SKILL-DISCO + ReActBackbone=GPT-4o2026.0692.48.7822,1006070.036
ReActBackbone=GPT-4o2026.068219.2935,6006260.045
AWMofflineBackbone=GPT-4o2026.0654.4811.3438,9082750.0484
ASIofflineBackbone=GPT-4o2026.0647.0111.4368,9452860.0533