Research map / Probabilistic and scientific modeling / Variational and generative sampling
Additive smoothing error in backward variational inference for general state-space models
Mathis Chagneux, Elisabeth Gassiat, Pierre Gloaguen, Sylvain Le Corff
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 |
|---|---|---|---|
| use a GRU backward parameterization instead of the conjugate backward model (ablation) | Backward variational smoother · Nonlinear emission SSM, additive smoothing error vs FFBSi | additive smoothing error | Improved |
| use a forward (two-filter) conjugate model instead of a backward factorization (ablation) | Variational smoother · Nonlinear emission SSM | growth of additive error with n | No clear effect |
| train GRU backward model on increasingly long subsequences instead of only the full sequence (ablation) | GRU Backward · Chaotic RNN with Student-t noise | additive error | Mixed |
| add per-timestep re-optimization of terminal laws (Campbell et al.) to conjugate models | Conjugate Forward and Backward models · Chaotic RNN with Student-t noise | additive error | Improved |
Closest papers by meaning
- Online Variational Filtering and Parameter Learning (NeurIPS 2021 · similarity 0.89)
- Efficient Online Variational Estimation via Monte Carlo Sampling (ICML 2026 · similarity 0.88)
- Self-Supervised Inference in State-Space Models (ICLR 2022 · similarity 0.88)
- Robust, Accurate Stochastic Optimization for Variational Inference (NeurIPS 2020 · similarity 0.87)
- Globally Convergent Variational Inference (NeurIPS 2024 · similarity 0.87)
- Online Variational Sequential Monte Carlo (ICML 2024 · 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.
