Agentic Reasoning on ALFWorld
76.02Success RateEAPO-8B
Evaluation Results
| Method | Links | |
|---|---|---|
| EAPO-8BModel Category=Ours2026.05 | 76.02 | |
| EAPO-8B2026.05 | 76.02 | |
| EAPO-4BModel Category=Ours2026.05 | 69 | |
| EAPO-4B2026.05 | 69 | |
| HTPOBase Model=Qwen3-8B2026.05 | 61.4 | |
| LAMERCategory=Training Method2026.05 | 61.26 | |
| DAPOBase Model=Qwen3-8B2026.05 | 60 | |
| GRPOBase Model=Qwen3-8B2026.05 | 59.3 | |
| EAPO-1.7B/2BModel Category=Ours2026.05 | 58.5 | |
| EAPO-1.7B/2B2026.05 | 58.5 | |
| GiGPOCategory=Training Method2026.05 | 56.76 | |
| Claude-4-Sonnet-0929Model Category=Closed-source Models2026.05 | 56.09 | |
| Claude-4-Sonnet-0929Category=Closed-source2026.05 | 56.09 | |
| DeepMiner-Mano-72BModel Category=Closed-source Models2026.05 | 54.12 | |
| DeepMiner-Mano-72BCategory=Closed-source2026.05 | 54.12 | |
| DAPOCategory=Training Method2026.05 | 51.86 | |
| UI-TARS-2-2509Model Category=Closed-source Models2026.05 | 51.21 | |
| UI-TARS-2-2509Category=Closed-source2026.05 | 51.21 | |
| Qwen3-VL-235B-A22BModel Category=Open-source Models2026.05 | 50.81 | |
| Qwen3-VL-235B-A22BCategory=Open-source2026.05 | 50.81 | |
| Qwen3-VL-32BModel Category=Open-source Models2026.05 | 49.69 | |
| Claude-3.7-SonnetModel Category=Closed-source Models2026.05 | 48.61 | |
| Seed1.5-VL-250717Model Category=Closed-source Models2026.05 | 47.53 | |
| Seed1.5-VL-250717Category=Closed-source2026.05 | 47.53 | |
| DeepMiner-Mano-7BModel Category=Closed-source Models2026.05 | 46.27 | |
| GRPOCategory=Training Method2026.05 | 46.13 | |
| DART-GUI-7BModel Category=Open-source Models2026.05 | 45.78 | |
| DART-GUI-7BCategory=Open-source2026.05 | 45.78 | |
| OpenAI CUAModel Category=Closed-source Models2026.05 | 44.88 | |
| UI-TARS-72B-dpoModel Category=Open-source Models2026.05 | 44.36 | |
| OpenAI CUA o3Model Category=Closed-source Models2026.05 | 42.31 | |
| OpenAI CUA o3Category=Closed-source2026.05 | 42.31 | |
| OpenCUA-32BModel Category=Open-source Models2026.05 | 39.99 | |
| OpenCUA-32BCategory=Open-source2026.05 | 39.99 | |
| GUI-Owl-7BModel Category=Open-source Models2026.05 | 38.47 | |
| GUI-Owl-7BCategory=Open-source2026.05 | 38.47 | |
| ARPOModel Category=Open-source Models2026.05 | 37.2 | |
| ARPOCategory=Open-source2026.05 | 37.2 | |
| TianXi-Action-7BModel Category=Closed-source Models2026.05 | 36.57 | |
| TianXi-Action-7BCategory=Closed-source2026.05 | 36.57 | |
| ZeroGUIModel Category=Open-source Models2026.05 | 35.76 | |
| ZeroGUICategory=Open-source2026.05 | 35.76 | |
| OpenCUA-7BModel Category=Open-source Models2026.05 | 34.69 | |
| UI-TARS-7BModel Category=Open-source Models2026.05 | 31.82 | |
| UI-TARS-7BCategory=Open-source2026.05 | 31.82 |