Instruction Following on ALFWorld (test)
91.4Success RateMemP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MemPModel=Qwen2.5-72B (Teacher)2026.06 | 91.4 | — | — | |
| No Mem.Model=Qwen2.5-72B (Teacher)2026.06 | 87.1 | — | — | |
| DuoMemModel=Qwen3-4B-Inst. (4B)2026.06 | 77.9 | 73.6 | 22.9 | |
| +LoRAModel=Qwen3-4B-Inst. (4B)2026.06 | 72.1 | — | — | |
| DuoMemModel=Gemma3-12B-it (12B)2026.06 | 66.4 | 52.1 | 25 | |
| MemPModel=Qwen3-8B (8B)2026.06 | 64.3 | — | — | |
| DuoMemModel=Qwen3-8B (8B)2026.06 | 64.3 | 23.6 | 0 | |
| +CDModel=Qwen3-8B (8B)2026.06 | 61.4 | — | — | |
| +LoRAModel=Qwen3-8B (8B)2026.06 | 60.7 | — | — | |
| +CDModel=Qwen3-4B-Inst. (4B)2026.06 | 56.4 | — | — | |
| +LoRAModel=Gemma3-12B-it (12B)2026.06 | 56.4 | — | — | |
| DuoMemModel=Gemma4-E2B-it (2B)2026.06 | 55.7 | 53.6 | 41.4 | |
| aTTTSignal=Self, Model=Qwen3.5-9B2026.07 | 55.7 | — | — | |
| aTTTSignal=Env, Model=Qwen3.5-9B2026.07 | 55.5 | — | — | |
| MemPModel=Qwen3-4B-Inst. (4B)2026.06 | 55 | — | — | |
| aTTTSignal=Summary, Model=Qwen3.5-9B2026.07 | 54.3 | — | — | |
| Sequence filterSignal=Summary, Model=Qwen3.5-9B2026.07 | 54 | — | — | |
| Sequence filterSignal=Self, Model=Qwen3.5-9B2026.07 | 53.8 | — | — | |
| No filterSignal=Self, Model=Qwen3.5-9B2026.07 | 53.1 | — | — | |
| qTTTModel=Qwen3.5-9B2026.07 | 52.4 | — | — | |
| Sequence filterSignal=Env, Model=Qwen3.5-9B2026.07 | 52.1 | — | — | |
| No filterSignal=Summary, Model=Qwen3.5-9B2026.07 | 52.1 | — | — | |
| No filterSignal=Env, Model=Qwen3.5-9B2026.07 | 51 | — | — | |
| ReActModel=Qwen3.5-9B2026.07 | 50.7 | — | — | |
| +LoRAModel=Gemma4-E2B-it (2B)2026.06 | 46.4 | — | — | |
| +CDModel=Gemma3-12B-it (12B)2026.06 | 46.4 | — | — | |
| MemPModel=Gemma3-12B-it (12B)2026.06 | 41.4 | — | — | |
| No Mem.Model=Qwen3-8B (8B)2026.06 | 40.7 | — | — | |
| aTTTSignal=Summary, Model=Gemma-3-12B2026.07 | 26.7 | — | — | |
| Sequence filterSignal=Summary, Model=Gemma-3-12B2026.07 | 26.2 | — | — | |
| No filterSignal=Summary, Model=Gemma-3-12B2026.07 | 25.5 | — | — | |
| aTTTSignal=Env, Model=Gemma-3-12B2026.07 | 25 | — | — | |
| aTTTSignal=Self, Model=Gemma-3-12B2026.07 | 24.3 | — | — | |
| Sequence filterSignal=Self, Model=Gemma-3-12B2026.07 | 23.8 | — | — | |
| Sequence filterSignal=Env, Model=Gemma-3-12B2026.07 | 22.9 | — | — | |
| qTTTModel=Gemma-3-12B2026.07 | 22.6 | — | — | |
| ReActModel=Gemma-3-12B2026.07 | 22.4 | — | — | |
| No filterSignal=Self, Model=Gemma-3-12B2026.07 | 22.1 | — | — | |
| No filterSignal=Env, Model=Gemma-3-12B2026.07 | 21.7 | — | — | |
| +CDModel=Gemma4-E2B-it (2B)2026.06 | 16.4 | — | — | |
| MemPModel=Gemma4-E2B-it (2B)2026.06 | 14.3 | — | — | |
| No Mem.Model=Gemma3-12B-it (12B)2026.06 | 14.3 | — | — | |
| Sequence filterSignal=Env, Model=Qwen3.5-4B2026.07 | 5.5 | — | — | |
| aTTTSignal=Summary, Model=Qwen3.5-4B2026.07 | 4.8 | — | — | |
| aTTTSignal=Env, Model=Qwen3.5-4B2026.07 | 4.5 | — | — | |
| No filterSignal=Summary, Model=Qwen3.5-4B2026.07 | 4.5 | — | — | |
| No Mem.Model=Qwen3-4B-Inst. (4B)2026.06 | 4.3 | — | — | |
| No filterSignal=Env, Model=Qwen3.5-4B2026.07 | 4.3 | — | — | |
| Sequence filterSignal=Summary, Model=Qwen3.5-4B2026.07 | 4.3 | — | — | |
| aTTTSignal=Self, Model=Qwen3.5-4B2026.07 | 3.6 | — | — | |
| Sequence filterSignal=Self, Model=Qwen3.5-4B2026.07 | 3.1 | — | — | |
| qTTTModel=Qwen3.5-4B2026.07 | 2.6 | — | — | |
| No Mem.Model=Gemma4-E2B-it (2B)2026.06 | 2.1 | — | — | |
| No filterSignal=Self, Model=Qwen3.5-4B2026.07 | 2.1 | — | — | |
| ReActModel=Qwen3.5-4B2026.07 | 1.9 | — | — |