Research map / Probabilistic and scientific modeling / Variational and generative sampling
Continuous Latent Process Flows
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 |
|---|---|---|---|
| add piecewise variational posterior (vs global posterior SDE, CLPF-Global) (ablation) | CLPF · synthetic GBM lambda=2 | test NLL: 0.447 -> 0.444 | No clear effect |
| add piecewise variational posterior (vs global posterior SDE, CLPF-Global) (ablation) | CLPF · synthetic LSDE lambda=2 | test NLL: -0.821 -> -0.831 | Helped |
| replace independent decoder with CTFP-style continuous decoder (ablation) | CLPF · synthetic SLC | test NLL: 7.924 -> -2.620 | Helped |
| replace independent decoder with CTFP-style continuous decoder (ablation) | CLPF · synthetic CAR lambda=2 | test NLL: 4.970 -> 1.322 | Helped |
| replace independent decoder with CTFP-style continuous decoder (ablation) | CLPF · synthetic GBM lambda=2 | test NLL: 0.800 -> 0.444 | Helped |
| use OU base process instead of Wiener process (ablation) | CLPF · synthetic SLC | test NLL: -1.885 -> -2.620 | Helped |
| use OU base process instead of Wiener process (ablation) | CLPF · synthetic CAR lambda=2 | test NLL: 1.041 -> 1.322 | Hurt |
| use OU base process instead of Wiener process (ablation) | CLPF · synthetic GBM lambda=2 | test NLL: 0.390 -> 0.444 | No clear effect |
| use piecewise posterior with CTFP decoding in latent SDE-like setting (CLPF vs Latent SDE) | CLPF · synthetic SLC lambda=20 | test NLL: 7.740 -> -2.620 | Helped |
| use CLPF instead of Latent CTFP | CLPF · synthetic CAR lambda=2 | test NLL: 249.839 -> 1.322 | Helped |
| use CLPF instead of VRNN on dense time grid (lambda=20) | CLPF · synthetic CAR lambda=20 | test NLL: 2.675 -> -0.077 | Helped |
| use VRNN instead of CLPF-ANODE for likelihood on real data | CLPF-ANODE · Mujoco-Hopper | test NLL: -14.694 -> -15.876 | Helped |
Closest papers by meaning
- Learning Coupled Continuous-Time Latent Dynamics from Irregular Events (ICML 2026 · similarity 0.88)
- Disentangling Continuous-Time Latent Dynamics: Identifiability of Latent SDEs via Diffusion Shifts (NeurIPS 2026 · similarity 0.86)
- ABC: Any-Subset Autoregression via Non-Markovian Diffusion Bridges in Continuous Time and Space (NeurIPS 2026 · similarity 0.85)
- Neural Stochastic Flows: Solver-Free Modelling and Inference for SDE Solutions (NeurIPS 2025 · similarity 0.85)
- Latent Representation and Simulation of Markov Processes via Time-Lagged Information Bottleneck (ICLR 2024 · similarity 0.85)
- Modeling Latent Non-Linear Dynamical System over Time Series (AAAI 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.