Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Driver-WM: A Driver-Centric Traffic-Conditioned Latent World Model for In-Cabin Dynamics Rollout
Haozhuang Chi, Daosheng Qiu, Hao Su, Haochen Liu, Zirui Li, Haoruo Zhang, Chen Lv
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 |
|---|---|---|---|
| add external out-cabin context stream (ablation) | Driver-WM · AIDE 5->5 rollout | TCR Macro-F1: 26.39 -> 90.15 | Improved |
| add external out-cabin context stream (ablation) | Driver-WM · AIDE 5->5 rollout | DER Macro-F1: 63.56 -> 72.61 | Improved |
| add external out-cabin context stream (ablation) | Driver-WM · AIDE 5->5 rollout | All MPJPE (px): 71.74 -> 71.47 | No clear effect |
| use attention transition core instead of RSSM/GRU (ablation) | Driver-WM · AIDE 5->5 rollout | All MPJPE (px): 78.42 -> 71.47 | Improved |
| add skeleton decoder pretraining (ablation) | Driver-WM · AIDE 5->5 rollout | All MPJPE (px): 71.47 -> 70.92 | No clear effect |
| add KL bottleneck (ablation) | Driver-WM · AIDE 5->5 rollout | All MPJPE (px): 71.47 -> 70.56 | No clear effect |
| add prompted external features (ablation) | Driver-WM · AIDE 5->5 rollout | All MPJPE (px): 76.54 -> 71.47 | Improved |
| add prompted external features (ablation) | Driver-WM · AIDE 5->5 rollout | TCR Macro-F1: 76.87 -> 90.15 | Improved |
| use causal rollout instead of non-causal bidirectional | Driver-WM · AIDE 5->5 rollout | All MPJPE (px): 72.15 -> 71.47 | No clear effect |
| add learned gated injection pathway versus disabling it (ablation) | Driver-WM · Test-time intervention | Delta All (px): 89.641 | Improved |
| force unmodulated injection (lambda_CA=1) (ablation) | Driver-WM · Test-time intervention | Delta All (px): 26.733 | Hurt |
| large temporal shift of external stream (ablation) | Driver-WM · Test-time intervention | Delta All (px): 0.793 | No clear effect |
Closest papers by meaning
- Driving in the Occupancy World: Vision-Centric 4D Occupancy Forecasting and Planning via World Models for Autonomous Driving (AAAI 2025 · similarity 0.87)
- DriveGPT: Scaling Autoregressive Behavior Models for Driving (ICML 2025 · similarity 0.86)
- Prioritizing Perception-Guided Self-Supervision: A New Paradigm for Causal Modeling in End-to-End Autonomous Driving (NeurIPS 2025 · similarity 0.86)
- NEST: A Neuromodulated Small-world Hypergraph Trajectory Prediction Model for Autonomous Driving (AAAI 2025 · similarity 0.86)
- BAT: Behavior-Aware Human-Like Trajectory Prediction for Autonomous Driving (AAAI 2024 · similarity 0.86)
- UniMotion: A Unified Motion Framework for Simulation, Prediction and Planning (NeurIPS 2025 · similarity 0.86)
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.
