Decision Making on ALFWorld Unseen v1
61.19Success Rate (SR)BeliefMem
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| BeliefMemBackbone=Qwen3-Next-80B-A3B-Instruct, Memory Corpus Size=50%2026.05 | 61.19 | 29.34 | |
| ReadAgentBackbone=Qwen3-Next-80B-A3B-Instruct2026.05 | 54.48 | 27.41 | |
| BeliefMemBackbone=Qwen3-Next-80B-A3B-Instruct, Memory Corpus Size=Full2026.05 | 53.75 | 30.49 | |
| Mem0Backbone=Qwen3-Next-80B-A3B-Instruct2026.05 | 41.04 | 33.16 | |
| MemoryBankBackbone=Qwen3-Next-80B-A3B-Instruct2026.05 | 38.06 | 34.99 | |
| LangMemBackbone=Qwen3-Next-80B-A3B-Instruct2026.05 | 31.34 | 37.17 | |
| A-MEMBackbone=Qwen3-Next-80B-A3B-Instruct2026.05 | 29.1 | 39.04 | |
| No-MemoryBackbone=Qwen3-Next-80B-A3B-Instruct2026.05 | 26.12 | 39.35 | |
| MemoryOSBackbone=Qwen3-Next-80B-A3B-Instruct2026.05 | 18.66 | 42.95 |