List your company

E-commerce Agentic Decision-making on WebShop

89.8ScoreG2PO
20.43238.44156.4574.459Jun 22, 2026

Evaluation results

MethodLinks
G2POType=RL Training, Backbone=Qwen2.5-7B-Instruct2026.0689.878.3
GiGPOType=RL Training, Backbone=Qwen2.5-7B-Instruct2026.0686.275.2
G2POType=RL Training, Backbone=Qwen2.5-1.5B-Instruct2026.0685.171.2
GiGPOType=RL Training, Backbone=Qwen2.5-1.5B-Instruct2026.0683.567.4
PPO (with critic)Type=RL Training, Backbone=Qwen2.5-7B-Instruct2026.0681.468.7
RLOOType=RL Training, Backbone=Qwen2.5-7B-Instruct2026.0680.365.7
GRPOType=RL Training, Backbone=Qwen2.5-7B-Instruct2026.0679.366.1
GRPOType=RL Training, Backbone=Qwen2.5-1.5B-Instruct2026.0675.856.8
RLOOType=RL Training, Backbone=Qwen2.5-1.5B-Instruct2026.0673.952.1
PPO (with critic)Type=RL Training, Backbone=Qwen2.5-1.5B-Instruct2026.0673.851.5
ReflexionType=Prompting, Backbone=Qwen2.5-7B-Instruct2026.0658.128.8
ReflexionType=Prompting, Backbone=Qwen2.5-1.5B-Instruct2026.0655.821.9
ReActType=Prompting, Backbone=Qwen2.5-7B-Instruct2026.0646.219.5
Gemini-2.5-ProType=Prompting, Backbone=Off-the-shelf2026.0642.535.9
ReActType=Prompting, Backbone=Qwen2.5-1.5B-Instruct2026.0640.111.3
DeepSeek-V3.2Type=Prompting, Backbone=Off-the-shelf2026.0631.618
Qwen3.5-397BType=Prompting, Backbone=Off-the-shelf2026.0627.617.2
Qwen2.5Type=Prompting, Backbone=Qwen2.5-7B-Instruct2026.0626.47.8
Qwen2.5Type=Prompting, Backbone=Qwen2.5-1.5B-Instruct2026.0623.15.2