Web Interaction on Mind2Web cross-Task
48.62Action Efficiency (EA)EDV
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| EDVBackbone Model=Ensemble2026.06 | 48.62 | 63.1 | 43.17 | 4.76 | |
| Reasoning Bank (RB)Backbone Model=Mimo-V2-Flash, Memory Strategy=RB2026.06 | 46.77 | 61.83 | 42.01 | 5.55 | |
| Reasoning Bank (RB)Backbone Model=GLM-4.7-FP8, Memory Strategy=RB2026.06 | 44.46 | 57.78 | 40.92 | 4.37 | |
| RouterBackbone Model=Ensemble2026.06 | 44.37 | 57.02 | 40.28 | 4.37 | |
| No Memory (NM)Backbone Model=GLM-4.7-FP8, Memory Strategy=NM2026.06 | 42.64 | 55.74 | 38.31 | 3.84 | |
| No Memory (NM)Backbone Model=Mimo-V2-Flash, Memory Strategy=NM2026.06 | 42.38 | 58.31 | 37.48 | 3.21 | |
| JudgeBackbone Model=Ensemble2026.06 | 41.37 | 56 | 35.14 | 3.06 | |
| No Memory (NM)Backbone Model=Minimax-M2.1, Memory Strategy=NM2026.06 | 39.92 | 59.14 | 34.19 | 3.57 | |
| Reasoning Bank (RB)Backbone Model=Minimax-M2.1, Memory Strategy=RB2026.06 | 39.06 | 59.75 | 33.95 | 3.17 |