List your company

Offline Action Consistency Diagnostic on AITW

74.2Action AccuracyQwen2.5-VL Agent + VisCritic
61.61664.88368.1571.417Jun 23, 2026

Evaluation results

MethodLinks
Qwen2.5-VL Agent + VisCriticbase_agent=Qwen2.5-VL Agent, critic_module=VisCritic2026.0674.2
Qwen2.5-VL Agent + GUI-Critic-R1base_agent=Qwen2.5-VL Agent, critic_module=GUI-Critic-R12026.0673.3
Qwen2.5-VL Agent + GuidNavbase_agent=Qwen2.5-VL Agent, critic_module=GuidNav2026.0673
Qwen2.5-VL Agent + GUI-PRAbase_agent=Qwen2.5-VL Agent, critic_module=GUI-PRA2026.0672.5
Qwen2.5-VL Agentbase_agent=Qwen2.5-VL Agent2026.0670.4
ShowUI + VisCriticbase_agent=ShowUI, critic_module=VisCritic2026.0669.8
ShowUI + GUI-PRAbase_agent=ShowUI, critic_module=GUI-PRA2026.0668.8
CogAgent + VisCriticbase_agent=CogAgent, critic_module=VisCritic2026.0667.1
ShowUIbase_agent=ShowUI2026.0666.7
SeeClick + GUI-Critic-R1base_agent=SeeClick, critic_module=GUI-Critic-R12026.0664.8
CogAgentbase_agent=CogAgent2026.0664.8
SeeClick + GUI-PRAbase_agent=SeeClick, critic_module=GUI-PRA2026.0664.5
SeeClick + VisCriticbase_agent=SeeClick, critic_module=VisCritic2026.0664.5
SeeClick + BacktrackAgentbase_agent=SeeClick, critic_module=BacktrackAgent2026.0664.2
SeeClickbase_agent=SeeClick2026.0662.1