Research map / Probabilistic and scientific modeling / Causal discovery
Learning Temporally Causal Latent Processes from General Temporal Data
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 causal process prior (ablation) | β-VAE on NP (Dense) · NP (Dense) | MCC: 0.446 ± 0.004 -> 0.721 ± 0.121 | Helped |
| add nonstationary flow (ablation) | LEAP + Causal Process Prior · NP (Dense) | MCC: 0.721 ± 0.121 -> 0.939 ± 0.008 | Helped |
| add noise discriminator (ablation) | LEAP + Nonstationary Flow · NP (Dense) | MCC: 0.939 ± 0.008 -> 0.983 ± 0.002 | Helped |
| add causal process prior (ablation) | β-VAE on VAR · VAR | MCC: 0.495 ± 0.007 -> 0.752 ± 0.035 | Helped |
| add nonstationary flow (ablation) | LEAP + Causal Process Prior · VAR | MCC: 0.752 ± 0.035 -> 0.935 ± 0.014 | Helped |
| add noise discriminator (ablation) | LEAP + Nonstationary Flow · VAR | MCC: 0.935 ± 0.014 -> 0.978 ± 0.004 | Helped |
| use history and nonstationarity information rather than ignoring both | LEAP · synthetic NP and VAR datasets | MCC | Helped |
| assume independent sources despite time-delayed causal relations | SlowVAE and PCL · synthetic NP and VAR datasets | MCC | Hurt |
| use linear causal transitions on nonlinear NP processes | LEAP-VAR · synthetic NP datasets | MCC | Helped |
| violate conditional independence with instantaneous causal relations | LEAP · synthetic VAR datasets | MCC | Hurt |
| use Gaussian noise rather than the required non-Gaussian noise | LEAP · synthetic VAR datasets | MCC | Hurt |
| reduce the number of regimes below the sufficient-variability requirement | LEAP · synthetic NP datasets | MCC | No clear effect |
Closest papers by meaning
- On the Identification of Temporal Causal Representation with Instantaneous Dependence (ICLR 2025 · similarity 0.91)
- Towards Identifiability of Hierarchical Temporal Causal Representation Learning (NeurIPS 2025 · similarity 0.89)
- Differentiable Causal Discovery for Latent Hierarchical Causal Models (ICLR 2025 · similarity 0.88)
- A Sparsity Principle for Partially Observable Causal Representation Learning (ICML 2024 · similarity 0.88)
- Causal Temporal Representation Learning with Nonstationary Sparse Transition (NeurIPS 2024 · similarity 0.88)
- CaRiNG: Learning Temporal Causal Representation under Non-Invertible Generation Process (ICML 2024 · similarity 0.88)
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.