List your company

Web-based Task Automation on WebShop (test)

85.5ScoreECPO
2.61224.13145.6567.169Jun 4, 2026

Evaluation results

MethodLinks
ECPOBase Model=Qwen2.5-7B-Instruct2026.0685.574.7
ECPOBase Model=Qwen2.5-1.5B-Instruct2026.0684.471.9
GiGPOw/stdBase Model=Qwen2.5-7B-Instruct2026.0684.472.4
PPOBase Model=Qwen2.5-7B-Instruct2026.0681.468.7
GiGPOw/stdBase Model=Qwen2.5-1.5B-Instruct2026.0680.464.6
RLOOBase Model=Qwen2.5-7B-Instruct2026.0680.365.7
GRPOBase Model=Qwen2.5-7B-Instruct2026.0679.366.1
RLOOBase Model=Qwen2.5-1.5B-Instruct2026.0673.952.1
PPOBase Model=Qwen2.5-1.5B-Instruct2026.0673.851.5
GRPOBase Model=Qwen2.5-1.5B-Instruct2026.0664.152.6
ReflexionBase Model=Qwen2.5-7B-Instruct2026.0658.128.8
ReflexionBase Model=Qwen2.5-1.5B-Instruct2026.0655.821.9
ReActBase Model=Qwen2.5-7B-Instruct2026.0646.219.5
ReActBase Model=Qwen2.5-1.5B-Instruct2026.0640.111.3
PromptingBase Model=Qwen2.5-7B-Instruct2026.0626.47.8
PromptingBase Model=Qwen2.5-1.5B-Instruct2026.0623.15.2
GLM-5.1Model Family=Closed-source LLMs2026.0612.98.1
Gemini-3.5-FlashModel Family=Closed-source LLMs2026.065.83.9