List your company

Web interaction on Mind2Web (cross-website)

44.79Action AccuracyEDV
6.476416.423226.3736.3168May 11, 2026May 18, 2026May 25, 2026Jun 1, 2026Jun 8, 2026Jun 15, 2026Jun 23, 2026

Evaluation results

MethodLinks
EDVBackbone Model=Ensemble2026.0644.794.7655.6436.56
Reasoning Bank (RB)Backbone Model=GLM-4.7-FP8, Memory Strategy=RB2026.0641.664.5253.4235.83
RouterBackbone Model=Ensemble2026.0639.913.3952.1531.76
No Memory (NM)Backbone Model=GLM-4.7-FP8, Memory Strategy=NM2026.0639.263.3951.6531.17
Reasoning Bank (RB)Backbone Model=Mimo-V2-Flash, Memory Strategy=RB2026.0636.814.1249.6331.78
Reasoning Bank (RB)Backbone Model=Minimax-M2.1, Memory Strategy=RB2026.0636.422.8249.7929.72
No Memory (NM)Backbone Model=Mimo-V2-Flash, Memory Strategy=NM2026.0635.554.7648.0230.29
Evolving-RLSkill injection=false2026.0535.153.39——
Evolving-RLSkill injection=true2026.0535.141.81——
No Memory (NM)Backbone Model=Minimax-M2.1, Memory Strategy=NM2026.0635.032.8249.5929.13
JudgeBackbone Model=Ensemble2026.0635.011.8746.8527.63
GRPOSkill injection=true2026.0526.20.9——
GRPOSkill injection=false2026.0524.610——
ReasoningBank2026.0517.760.56——
Memento2026.0515.460.56——
Base ModelSkill injection=true2026.0515.010——
Base ModelSkill injection=false2026.057.950——