Research map / Probabilistic and scientific modeling / Dynamical systems
Hidden Parameter Recurrent State Space Models For Changing Dynamics Scenarios
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| condition recurrent state dynamics on a latent task parameter inferred from context | RSSM · Pneumatic muscular arm, No Imputation | RMSE (10^-3): 3.088\pm 0.046 -> 2.30 \pm0.043 | Helped |
| condition recurrent state dynamics on a latent task parameter inferred from context | RSSM · Pneumatic muscular arm, 50 % Imputation | RMSE (10^-3): 3.223\pm0.014 -> 2.47\pm 0.012 | Helped |
| condition recurrent state dynamics on a latent task parameter inferred from context | RSSM · Franka, 50 % Imputation | RMSE (10^-4): 3.398 \pm 0.062 -> 2.843 \pm 0.024 | Helped |
| condition recurrent state dynamics on a latent task parameter inferred from context | RSSM · Franka, No Imputation | RMSE (10^-4): 2.765 \pm0.004 vs 2.833 \pm 0.024 | Mixed |
| condition recurrent state dynamics on a latent task parameter inferred from context | RSSM · Mobile Robot, No Imputation | RMSE (10^-5): 4.57\pm0.013 -> 2.96 \pm 0.212 | Helped |
| replace the permutation-invariant set context encoder with a recurrent encoder (ablation) | HiP-RSSM · Mobile Robot, No Imputation | RMSE (10^-5): 2.96 \pm 0.212 -> 5.10 \pm0.041 | Hurt |
| replace the permutation-invariant set context encoder with a recurrent encoder (ablation) | HiP-RSSM · Mobile Robot, 50 % Imputation | RMSE (10^-5): 6.15 \pm 0.327 -> 10.12\pm 0.112 | Hurt |
| replace the permutation-invariant set context encoder with a recurrent encoder (ablation) | HiP-RSSM · Mobile Robot | Training time per epoch: 6.71 -> 14.13 | Hurt |
| use a nonlinear feed-forward task transformation instead of linear or locally linear task transformation (ablation) | HiP-RSSM · robotic dynamics benchmarks | prediction RMSE | Helped |
| train with RMSE rather than Gaussian log-likelihood (ablation) | HiP-RSSM · robotic dynamics benchmarks | prediction RMSE | Helped |
| use latent task-conditioned dynamics rather than soft switching between dynamics matrices | HiP-RSSM · Wheeled Mobile Robot | multi-step prediction RMSE | Helped |
Closest papers by meaning
- MaRK: Markov-adapted Recurrent Kernels for Dynamic Operator Conditioning in State Space Models (NeurIPS 2026 · similarity 0.86)
- UnHiPPO: Uncertainty-aware Initialization for State Space Models (ICML 2025 · similarity 0.86)
- Understanding the Differences in Foundation Models: Attention, State Space Models, and Recurrent Neural Networks (NeurIPS 2024 · similarity 0.86)
- Learning long range dependencies through time reversal symmetry breaking (NeurIPS 2025 · similarity 0.85)
- A Generalizable Physics-Enhanced State Space Model for Long-Term Dynamics Forecasting in Complex Environments (ICML 2025 · similarity 0.85)
- Identifying latent state transitions in non-linear dynamical systems (ICLR 2025 · similarity 0.85)
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.