Research map / 3D vision / Face and shape
A Resolution-Agnostic Geometric Transformer for Chromosome Modeling Using Inertial Frame
Yize Zhou, Haorui Li, Shengchao Liu
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 inertial frame alignment (ablation) | IG-3DMAX · Frontal cortex 320KB | dRMSE: 0.1641 -> 0.1547 | Improved |
| add inertial frame alignment (ablation) | IG-3DMAX · Frontal cortex 40KB | dRMSE: 0.2385 -> 0.2382 | No clear effect |
| add 3D RoPE (ablation) | IG-3DMAX · Frontal cortex 40KB | dRMSE: 0.2454 -> 0.2382 | Improved |
| add Nystrom positional encoding (ablation) | IG-3DMAX · Frontal cortex 40KB | dRMSE: 0.2496 -> 0.2382 | Improved |
| use structural loss only (alpha=1.0) instead of mixed loss (ablation) | IG-3DMAX · Frontal cortex 40KB | dRMSE: 0.2407 -> 0.2788 | Hurt |
| use structural loss only (alpha=1.0) instead of mixed loss (ablation) | IG-3DMAX · Frontal cortex 320KB | dRMSE: 0.1696 -> 0.1453 | Improved |
| drop structural loss (alpha=0) (ablation) | IG-3DMAX · Frontal cortex 40KB | dRMSE: 0.2407 -> 0.2410 | No clear effect |
| apply InertialGenome to 3DMAX coordinates | 3DMAX · Frontal cortex 320KB | dSCC: 0.2780 -> 0.9006 | Improved |
| apply InertialGenome to 3DMAX coordinates | 3DMAX · B-Lymphocyte 100KB | dSCC: 0.6548 -> 0.8708 | Improved |
| use IG-3DMAX instead of HiCEGNN | HiCEGNN · Frontal cortex 40KB | dSCC: 0.2506 -> 0.7187 | Improved |
| use Gram-matrix coordinates as input instead of 3DMAX | InertialGenome · Frontal cortex 40KB | dSCC: 0.7187 -> 0.0083 | Hurt |
| transfer from 320kb to 40kb | IG-3DMAX · Cross-resolution transfer | dSCC: 0.6132 -> 0.6528 | Improved |
Closest papers by meaning
- Multimodal 3D Genome Pre-training (NeurIPS 2025 · similarity 0.81)
- BOE-ViT: Boosting Orientation Estimation with Equivariance in Self-Supervised 3D Subtomogram Alignment (CVPR 2025 · similarity 0.81)
- Beyond Sequence: Impact of Geometric Context for RNA Property Prediction (ICLR 2025 · similarity 0.80)
- Enforcing Latent Euclidean Geometry in Single-Cell VAEs for Manifold Interpolation (ICML 2025 · similarity 0.80)
- SILSA: Sliding-Window Slice Latents for Topology-Preserving High-Resolution 3D Generation (NeurIPS 2026 · similarity 0.80)
- Hyperbolic Genome Embeddings (ICLR 2025 · similarity 0.80)
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.
