List your company
ResearchFormerly WizWand

Web Task Completion on MiniWoB++ Held-out Goals (test)

70.5Success RateClaude-3.5-Sonnet
27.8638.935061.07Jul 5, 2025
Updated 8mo ago

Evaluation results

MethodLinks
Claude-3.5-SonnetBackbone Model=Claude-3.5-Sonnet, Training Protocol=Instruct, Model Role=Baseline2025.0770.5
o1-MiniBackbone Model=o1-Mini, Training Protocol=Instruct, Model Role=Baseline2025.0769.7
Llama-3.1-8B SFT+RL (Ours)Backbone Model=Llama-3.1-8B, Training Protocol=SFT+RL, Model Role=Agent2025.0767.2
Llama-3.1-405B-InstructBackbone Model=Llama-3.1-405B, Training Protocol=Instruct, Model Role=Baseline2025.0765.9
GPT-4oBackbone Model=GPT-4o, Training Protocol=Instruct, Model Role=Baseline2025.0765.7
Qwen-2.5-7B SFT+RL (Ours)Backbone Model=Qwen-2.5-7B, Training Protocol=SFT+RL, Model Role=Agent2025.0765
Llama-3.3-70B Instruct (Teacher)Backbone Model=Llama-3.3-70B, Training Protocol=Instruct, Model Role=Teacher2025.0763.2
Qwen-2.5-72B Instruct (Teacher)Backbone Model=Qwen-2.5-72B, Training Protocol=Instruct, Model Role=Teacher2025.0761
Llama-3.1-70B-InstructBackbone Model=Llama-3.1-70B, Training Protocol=Instruct, Model Role=Baseline2025.0757
Qwen-2.5-7B SFT (Ours)Backbone Model=Qwen-2.5-7B, Training Protocol=SFT, Model Role=Agent2025.0757
GPT-4o-MiniBackbone Model=GPT-4o-Mini, Training Protocol=Instruct, Model Role=Baseline2025.0756.2
Llama-3.1-8B SFT (Ours)Backbone Model=Llama-3.1-8B, Training Protocol=SFT, Model Role=Agent2025.0755.2
Qwen-2.5-7B RL (Ours)Backbone Model=Qwen-2.5-7B, Training Protocol=RL, Model Role=Agent2025.0752.5
Llama-3.1-8B RL (Ours)Backbone Model=Llama-3.1-8B, Training Protocol=RL, Model Role=Agent2025.0743.5
Qwen-2.5-7B Instruct (Student)Backbone Model=Qwen-2.5-7B, Training Protocol=Instruct, Model Role=Student2025.0732.8
Llama-3.1-8B Instruct (Student)Backbone Model=Llama-3.1-8B, Training Protocol=Instruct, Model Role=Student2025.0729.5