Inventory Management on Inventory Management decreasing-diverse GPT-5
31.33Relative Gap (Δ)InvAgent
Evaluation results
| Method | Links | ||
|---|---|---|---|
| InvAgentBackbone=GPT-5, Reasoning effort=medium, Strategy=w/o strategy2026.02 | 31.33 | 0 | |
| AIM-RMBackbone=GPT-5, Reasoning effort=medium, Strategy=w/ RL log2026.02 | 51.81 | 0 | |
| AIM-RMBackbone=GPT-5, Reasoning effort=medium, Strategy=w/o RL log2026.02 | 55.18 | 3.8 | |
| InvAgentBackbone=GPT-5, Reasoning effort=medium, Strategy=w/ step desc and our strategy2026.02 | 60.84 | 0 | |
| InvAgentBackbone=GPT-5, Reasoning effort=medium, Strategy=w/ step desc2026.02 | 65.06 | 0 | |
| InvAgentBackbone=GPT-5, Reasoning effort=medium, Strategy=w/ strategy2026.02 | 131.33 | 0 |