Research map / Probabilistic and scientific modeling / Dynamical systems
Deep Generative Markov State Models
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 |
|---|---|---|---|
| Use smooth probabilistic state memberships instead of hard cluster assignments | DeepResampleMSM versus standard MSM · Prinz potential | relaxation timescales | Helped |
| Increase the standard MSM state count from four to ten | standard MSM · Prinz potential | relaxation timescales | Helped |
| Use ML-ED training instead of ED alone | DeepGenMSM · Prinz potential | stationary distribution and kinetics | Helped |
| Train a generative model rather than resampling empirical configurations | DeepGenMSM · Prinz potential | stationary distribution and kinetics | Mixed |
| Use six-state DeepResampleMSM instead of six-state hard-cluster MSM | Alanine dipeptide · Alanine dipeptide | relaxation timescales | Helped |
| Increase the standard MSM state count from six to 100 | standard MSM · Alanine dipeptide | relaxation timescales | Helped |
| Train the generator with ML-ED using ML-trained state memberships | DeepGenMSM · Alanine dipeptide | kinetics | Helped |
| Generate configurations recursively with DeepGenMSM | DeepGenMSM · Alanine dipeptide | stationary distribution | Mixed |
| Remove training data from one metastable state | DeepGenMSM · Alanine dipeptide, held-out metastable states | quality of generated configurations | Mixed |
| Vary network depth, width, and random-input dimension (ablation) | DeepGenMSM generator · Prinz potential | KL divergence of generated stationary distribution | No clear effect |
Closest papers by meaning
- Generative Modeling of Molecular Dynamics Trajectories (NeurIPS 2024 · similarity 0.91)
- Foundation Inference Models for Markov Jump Processes (NeurIPS 2024 · similarity 0.87)
- Atomic Trajectory Modeling with State Space Models for Biomolecular Dynamics (NeurIPS 2026 · similarity 0.86)
- Unlocking Guidance for Discrete State-Space Diffusion and Flow Models (ICLR 2025 · similarity 0.85)
- eXponential FAmily Dynamical Systems (XFADS): Large-scale nonlinear Gaussian state-space modeling (NeurIPS 2024 · similarity 0.85)
- Action-Minimization Meets Generative Modeling: Efficient Transition Path Sampling with the Onsager-Machlup Functional (ICML 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.