List your company

Interactive Decision-making on AgentGym WebShop, ALFWorld, and BabyAI (test)

78.8WebShop Success RateSTRIDE
16.50432.67748.8565.023Jun 14, 2026

Evaluation results

MethodLinks
STRIDEType=Token-level RL, Base Model=Qwen2.5-VL-7B2026.0678.889.684.584.3
HET-DAPOType=Token-level RL, Base Model=Qwen2.5-VL-7B2026.0677.188.48382.8
AgentSTaRType=Offline RL, Base Model=Qwen2.5-VL-7B2026.0676.588.682.782.6
DPOType=Offline RL, Base Model=Qwen2.5-VL-7B2026.0675.386.578.380
AgentTraj-L-SFTType=SFT, Base Model=Qwen2.5-VL-7B2026.0673.583.274.277
PPOType=Online RL, Base Model=Qwen2.5-VL-7B2026.0668.583.569.873.9
RWRType=Offline RL, Base Model=Qwen2.5-VL-7B2026.0668.276.582.175.6
LLM-PlannerType=Prompting, Base Model=Qwen2.5-VL-7B2026.0618.968.582.556.6