Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
WestWorld: A Knowledge-Encoded Scalable Trajectory World Model for Diverse Robotic Systems
Yuchen Wang, Jiangtao Kong, Sizhe Wei, Xiaochang Li, Haohong Lin, Hongjue Zhao, Tianyi Zhou, Lu Gan, Huajie Shao
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 Sys-MoE layer instead of dense SSM (ablation) | WestWorld · Walker2D zero-shot | MAE (x1e-2): 18.707 -> 16.350 | Improved |
| add Sys-MoE layer instead of dense SSM (ablation) | WestWorld · Hopper zero-shot | MAE (x1e-2): 15.978 -> 13.731 | Improved |
| add Sys-MoE layer instead of dense SSM (ablation) | WestWorld · Franka zero-shot | MAE (x1e-2): 9.392 -> 7.737 | Improved |
| add structural embedding (ablation) | WestWorld · Walker2D zero-shot | MAE (x1e-2): 21.156 -> 16.350 | Improved |
| add structural embedding (ablation) | WestWorld · Hopper zero-shot | MAE (x1e-2): 16.227 -> 13.731 | Improved |
| add structural embedding (ablation) | WestWorld · Franka zero-shot | MAE (x1e-2): 7.897 -> 7.737 | Improved |
| pretrain before MPPI control training | WestWorld · Walker2d MPPI control | accumulated reward: 707.61 -> 2134.60 | Improved |
| pretrain before MPPI control training | WestWorld · Hopper MPPI control | accumulated reward: 554.92 -> 2253.51 | Improved |
| pretrain before MPPI control training | WestWorld · Go1 MPPI control | accumulated reward: 0.43 -> 2.20 | Improved |
| pretrain before MPPI control training | TDM · Hopper MPPI control | accumulated reward: 242.34 -> 154.64 | Hurt |
| pretrain before MPPI control training | TrajWorld · Walker2d MPPI control | accumulated reward: 395.23 -> 1933.52 | Improved |
| increase discretization bins from 128 to 256 (ablation) | WestWorld · Walker | MAE (x1e-2): 1.455 -> 1.415 | Improved |
Closest papers by meaning
- Trajectory World Models for Heterogeneous Environments (ICML 2025 · similarity 0.88)
- Tra-MoE: Learning Trajectory Prediction Model from Multiple Domains for Adaptive Policy Conditioning (CVPR 2025 · similarity 0.87)
- World Motion Models: Flexible Sequence Modeling of SE(3) Trajectories (NeurIPS 2026 · similarity 0.87)
- World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetry (NeurIPS 2026 · similarity 0.86)
- Predictive Inverse Dynamics Models are Scalable Learners for Robotic Manipulation (ICLR 2025 · similarity 0.86)
- DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning (ICML 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.
