| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Multi-agent coordination | Navigation ablation Clean (test) | Task Metric Value2.43 | 24 | |
| Multi-agent Reinforcement Learning | Navigation | Reward1.38 | 24 | |
| Multi-agent Observation Sharing | Navigation | Success Rate79 | 24 | |
| Rate of successful action changes | Navigation RIAL (test) | Rate of Successful Action Changes32 | 24 | |
| Navigation | Navigation | Task Metric Value2.89 | 24 | |
| Embodied Navigation | Navigation | Base Score88 | 17 | |
| Robot Navigation | Navigation | Average Total Discounted Reward11.9 | 16 | |
| Social Navigation | 500 random navigation (test) | SR97.2 | 10 | |
| End-to-end translation | Navigation | Accuracy99.1 | 10 | |
| Goal-driven Navigation | Navigation Average ObjNav & InsImageNav | Success Rate (SR)48.6 | 9 | |
| Reasoning | Navigation | Success Rate100 | 6 | |
| LLM Recovery | Navigation (Official) | Success Rate100 | 6 | |
| 2D Navigation | Navigation v3 | AP-3.25 | 6 | |
| 2D Navigation | Navigation v2 | AP-7.72 | 6 | |
| 2D Navigation | Navigation v1 | Average Precision-6.74 | 6 | |
| RNN-based navigation policy verification | Navigation 4x4 environment | Avg. Violation Rate1.42 | 5 | |
| Constrained Reinforcement Learning | Navigation | Episodic Reward217.6 | 5 | |
| LLM Recovery | Navigation Commitment-sensitive | Success Rate100 | 4 | |
| Multi-agent Navigation | Navigation (test) | Reward81 | 4 | |
| RNN-based navigation policy verification | Navigation (Nav) 8x8 environment | Average Violation Rate13.01 | 4 | |
| Drone Navigation | Navigation Synthetic | Gate Completion95 | 4 | |
| Drone Navigation | Navigation Non-Synthetic | Gate Score80 | 4 | |
| Navigation | Navigation Commit-sensitive | Success Rate100 | 3 | |
| Navigation | Navigation Entry-aligned | Success Rate100 | 3 | |
| System Safety and Recovery Audit | Navigation Commit-sensitive | Success Rate100 | 3 |