List your company

life-span-shortest-lived on ScienceWorld 20 unseen variants (test)

100ReturnEMPO2
2.34427.69753.0578.403Feb 26, 2026

Evaluation results

MethodLinks
EMPO2Backbone=Qwen2.5-7B-Instruct, Learning Strategy=Online RL, Memory usage at test time=None2026.02100
RetrospexBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Offline RL, Memory usage at test time=None2026.0267
NaiveBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Naive, Memory usage at test time=None2026.0256.8
GRPOBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Online RL, Memory usage at test time=None2026.0220.6
ReflexionBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Non-Parametric, Memory usage at test time=None2026.026.1