List your company

Embodied Task Completion on ALFWorld v1.0 (test)

100Pick Success RateReAct
66.30475.05283.892.548May 11, 2026

Evaluation results

MethodLinks
ReActMethod Category=Agent- or memory-based, Retrieved external skills=false2026.0510033.32553.823.826.750.5
ExpeLMethod Category=Agent- or memory-based, Retrieved external skills=false2026.0597.162.538.513.321.761.953.5
Zero-ShotMethod Category=Prompt-based, Retrieved external skills=true2026.0594.34081.238.517.45563.3
Few-ShotMethod Category=Prompt-based, Retrieved external skills=true2026.0594.36084.438.534.83064.1
ReflexionMethod Category=Agent- or memory-based, Retrieved external skills=false2026.0593.930522047.868.259.4
Mem0Method Category=Agent- or memory-based, Retrieved external skills=false2026.0593.9303226.717.418.242.2
Skill0Method Category=RL-based, Retrieved external skills=false2026.0593.687.559.866.757.678.474.2
SLIMMethod Category=RL-based, Retrieved external skills=true2026.0592.910091.478.388.581.287.5
SLIMMethod Category=RL-based, Retrieved external skills=false2026.0591.48046.961.573.98572.7
SkillRLMethod Category=RL-based, Retrieved external skills=true2026.0590.67554.676.267.787.575
GRPOMethod Category=RL-based, Retrieved external skills=false2026.0585.410049.864.653.154.267.2
GRPOMethod Category=RL-based, Retrieved external skills=true2026.0584.662.548.968.86180.868.8
Zero-ShotMethod Category=Prompt-based, Retrieved external skills=false2026.0582.9031.223.121.73041.4
Few-ShotMethod Category=Prompt-based, Retrieved external skills=false2026.05804037.57.717.41539.1
EvolveRMethod Category=RL-based, Retrieved external skills=false2026.0567.637.549.315.63636.539.8