List your company

Agent Behavior Adaptation on WebShop (WS) (test)

36.7Loop RatioQwen3-4B-Instruct
-1.4688.44118.3528.259Feb 2, 2026

Evaluation results

MethodLinks
Qwen3-4B-InstructModel Type=Non-thinking Model2026.0236.7
Phi-4-reasoningModel Type=Thinking Model2026.0235.7
Mistral-7B-InstructModel Type=Non-thinking Model2026.0217.5
gpt-oss-120bModel Type=Thinking Model2026.029.2
Qwen3-30B-A3B-InstructModel Type=Non-thinking Model2026.025.7
Qwen3-4B-ThinkingModel Type=Thinking Model2026.024.5
Ministral-3-14B-InstructModel Type=Non-thinking Model2026.024
Phi-4Model Type=Non-thinking Model2026.022.6
Llama-3.1-8B-InstructModel Type=Non-thinking Model2026.022.1
Llama-3.3-70B-InstructModel Type=Non-thinking Model2026.021.6
Glm-4-9B-ChatModel Type=Non-thinking Model2026.021.5
Qwen3-30B-A3B-ThinkingModel Type=Thinking Model2026.020.7
Gemini 2.5 FlashModel Type=Non-thinking Model2026.020.2
Glm-4-32B-0414Model Type=Non-thinking Model2026.020.1
DeepSeek-R1Model Type=Thinking Model2026.020.1
DeepSeek-V3.2Model Type=Non-thinking Model2026.020
Gemini 2.5 ProModel Type=Thinking Model2026.020