Research map / Probabilistic and scientific modeling / Causal discovery
On Causally Disentangled Representations
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 bounding-box-weighted reconstruction loss to SS-Factor-VAE | CANDLE · CANDLE | UC: 0.24 -> 0.27 | Helped |
| Add bounding-box-weighted reconstruction loss to SS-Factor-VAE | CANDLE · CANDLE | CG: 0.28 -> 0.28 | No clear effect |
| Use ρ=7 rather than ρ=5 latent dimensions per generative factor (ablation) | SS-FVAE-BB · CANDLE | UC: 0.27 -> 0.18 | Hurt |
| Use ρ=7 rather than ρ=5 latent dimensions per generative factor (ablation) | SS-FVAE-BB · CANDLE | CG: 0.28 -> 0.28 | No clear effect |
| Increase the number of attributed latent dimensions from ρ=1 to ρ=2 (ablation) | models evaluated on dSprites · dSprites | UC: 0.90 -> 0.67 | Hurt |
| Train on data with confounding | disentanglement models evaluated on dSprites · dSprites with confounding | UC | Hurt |
Closest papers by meaning
- C-Disentanglement: Discovering Causally-Independent Generative Factors under an Inductive Bias of Confounder (NeurIPS 2023 · similarity 0.93)
- Disentangled Representation Learning in Non-Markovian Causal Systems (NeurIPS 2024 · similarity 0.93)
- Toward Causal Generative Modeling: From Representation to Generation (AAAI 2025 · similarity 0.91)
- Synergy Between Sufficient Changes and Sparse Mixing Procedure for Disentangled Representation Learning (ICLR 2025 · similarity 0.90)
- Transferring disentangled representations: bridging the gap between synthetic and real images (NeurIPS 2024 · similarity 0.89)
- $\alpha$TC-VAE: On the relationship between Disentanglement and Diversity (ICLR 2024 · similarity 0.89)
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.