List your company

chemistry-mix-paint-tertiary-color on ScienceWorld 20 unseen variants (test)

42.6ReturnGRPO
6.215.6525.134.55Feb 26, 2026

Evaluation results

MethodLinks
GRPOBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Online RL, Memory usage at test time=None2026.0242.6
EMPO2Backbone=Qwen2.5-7B-Instruct, Learning Strategy=Online RL, Memory usage at test time=None2026.0239.2
ReflexionBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Non-Parametric, Memory usage at test time=None2026.0236.9
NaiveBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Naive, Memory usage at test time=None2026.0233.9
RetrospexBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Offline RL, Memory usage at test time=None2026.027.6