Research map / Probabilistic and scientific modeling / Variational and generative sampling
Orthogonality-Enforced Latent Space in Autoencoders: An Approach to Learning 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 |
|---|---|---|---|
| Euler encoding and normalization layer (full DAE) vs. without them (ablation) | DAE · disentanglement scores (Table 14) | disentanglement scores | Helped |
| DAE vs. other models on XY to XY CS (adding color and shape) | DAE · XY vs XY CS | jemmig/dcimig/dci-rf percentage change: largest drop 17.2% vs 39.7%-92.5% for others | Helped |
| Gaussian interpolation / no regularizer in AE | AE · GF-Score across datasets | GF-Score | Hurt |
| Proposed DAE vs. baselines on 3D Airplane, Sprites, 2D Toy | DAE · 2D Toy, 3D Airplane, Sprites | all disentanglement metrics | Helped |
Closest papers by meaning
- Disentanglement via Latent Quantization (NeurIPS 2023 · similarity 0.94)
- Disentangling Disentanglement in Variational Autoencoders (ICML 2019 · similarity 0.93)
- Multifactor Sequential Disentanglement via Structured Koopman Autoencoders (ICLR 2023 · similarity 0.92)
- Tripod: Three Complementary Inductive Biases for Disentangled Representation Learning (ICML 2024 · similarity 0.91)
- Factorized Diffusion Autoencoder for Unsupervised Disentangled Representation Learning (AAAI 2024 · similarity 0.89)
- Disentangled Representation Learning via Flow Matching (NeurIPS 2026 · 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.