Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Solving Motion Planning Tasks with a Scalable Generative Model
Yihan Hu, Siqi Chai, Zhening Yang, Jingyu Qian, Kun Li, Wenxin Shao, Haichao Zhang, Wei Xu, Qiang Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use GUMP-m generative model for scene generation instead of TrafficGen | scene generation · WOD Motion scene generation (MMD) | Speed MMD: 0.1602 -> 0.0961 | Improved |
| use GUMP-m as world simulator instead of Trajeglish | sim agents · WOD Sim Agents test | minADE: 1.615 -> 1.590 | Improved |
| GUMP-m versus ensemble MVTE on sim agents meta metric | sim agents · WOD Sim Agents test | Meta Metrics: 0.6448 vs 0.6432 | No clear effect |
| imitate planning by iteratively unrolling ego states of GUMP-m compared with CKS-124m | open-loop planning · nuPlan val14 open-loop | Score: 88.0 -> 88.6 | Improved |
| GUMP-m versus CKS-124m on 8sADE | open-loop planning · nuPlan val14 open-loop | 8sADE: 1.777 vs 1.820 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- Trajeglish — Trajeglish: Traffic Modeling as Next-Token Prediction (ICLR 2024) · better on 1 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Scenario Dreamer: Vectorized Latent Diffusion for Generating Driving Simulation Environments (CVPR 2025 · similarity 0.91)
- Generating Multimodal Driving Scenes via Next-Scene Prediction (CVPR 2025 · similarity 0.90)
- SMART: Scalable Multi-agent Real-time Motion Generation via Next-token Prediction (NeurIPS 2024 · similarity 0.89)
- Generative Scenario Rollouts for End-to-End Autonomous Driving (NeurIPS 2026 · similarity 0.89)
- Generating Traffic Scenarios via In-Context Learning to Learn Better Motion Planner (AAAI 2025 · similarity 0.89)
- SceneDiffuser++: City-Scale Traffic Simulation via a Generative World Model (CVPR 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
