Research map / Probabilistic and scientific modeling / Molecules and proteins
Equivariant Latent Alignment via Flow Matching under Group Symmetries
Sunghyun Kim, Jaehoon Hahm, Jeongwoo Shin, Joonseok Lee
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 |
|---|---|---|---|
| add Residual Latent Flow correction to analytic latent rotation | NFT · RotatedMNIST (SO(2)) | PSNR: 28.21 -> 29.02 | Improved |
| add Residual Latent Flow correction | NFT · SmallNORB (SO(3), OOD) | PSNR: 23.13 -> 23.28 | Improved |
| add Residual Latent Flow correction | NFT · ABO-Material Day-to-Night (test) | PSNR: 22.73 -> 24.32 | Improved |
| add Residual Latent Flow correction | NFT · ComplexBRDFs (test) | PSNR: 16.04 -> 17.38 | Improved |
| add Residual Latent Flow correction | NFT · ModelNet10-SO(3) (OOD) | Angle Err.: 0.1746 -> 0.0430 | Improved |
| add Residual Latent Flow correction before rotation angle regression | NFT · ABO Day-to-Night rotation estimation | 1-cos angular discrepancy: 0.002507 -> 0.001131 (test); 0.004473 -> 0.001961 (OOD) | Improved |
| increase stochasticity sigma of the interpolant from 0 to 0.1 (ablation) | Residual Latent Flow · ABO-Material Day-to-Night (OOD) | PSNR: 22.84 -> 19.30 | Hurt |
| increase stochasticity sigma from 0 to 0.01 (ablation) | Residual Latent Flow · ABO-Material Day-to-Night (OOD) | Prediction Error: 0.0065 -> 0.0067 | Hurt |
| add flow correction at latent channel C=32 (ablation) | NFT · ABO Day-to-Night scalability (L=8) | PSNR: 20.20 -> 20.56 | Improved |
| add flow correction at latent channel C=128 (ablation) | NFT · ABO Day-to-Night scalability (L=8) | PSNR: 22.73 -> 24.32 | Improved |
| add flow correction at maximum degree L=2 (ablation) | NFT · ABO Day-to-Night scalability (C=128) | PSNR: 22.34 -> 22.73 | Improved |
| add flow correction at maximum degree L=8 (ablation) | NFT · ABO Day-to-Night scalability (C=128) | PSNR: 22.73 -> 24.32 | Improved |
Closest papers by meaning
- Probing Equivariance and Symmetry Breaking in Convolutional Networks (NeurIPS 2025 · similarity 0.88)
- Equivariance by Contrast: Identifiable Equivariant Embeddings from Unlabeled Finite Group Actions (NeurIPS 2025 · similarity 0.87)
- Neural Isometries: Taming Transformations for Equivariant ML (NeurIPS 2024 · similarity 0.87)
- Interpreting Equivariant Representations (ICML 2024 · similarity 0.87)
- Flow Equivariant Recurrent Neural Networks (NeurIPS 2025 · similarity 0.86)
- Variational Partial Group Convolutions for Input-Aware Partial Equivariance of Rotations and Color-Shifts (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.
