chemistry-mix on ScienceWorld 20 unseen variants (test)
42.7ReturnEMPO2
Evaluation Results
| Method | Links | |
|---|---|---|
| EMPO2Backbone=Qwen2.5-7B-Instruct, Learning Strategy=Online RL, Memory usage at test time=None2026.02 | 42.7 | |
| NaiveBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Naive, Memory usage at test time=None2026.02 | 42 | |
| RetrospexBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Offline RL, Memory usage at test time=None2026.02 | 20.8 | |
| GRPOBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Online RL, Memory usage at test time=None2026.02 | 12.4 | |
| ReflexionBackbone=Qwen2.5-7B-Instruct, Learning Strategy=Non-Parametric, Memory usage at test time=None2026.02 | 1.2 |