Research map / Architectures and efficiency / Attention and transformers
HiST: A Hierarchical Sparse Transformer for Cross-Modal Spatial Transcriptomics Modeling
Weiyi Wu, Xinwen Xu, Xingjian Diao, Siting Li, Zhi Wei, Alma Andersson, Jiang Gui
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 coordinate augmentation (ablation) | HiST · Brain (NCBI Geo) | R2: 0.149 -> 0.228 | Improved |
| add random rotate only (vs none) (ablation) | HiST · Brain (NCBI Geo) | R2: 0.149 -> 0.176 | Improved |
| add random drop only (vs none) (ablation) | HiST · Brain (NCBI Geo) | R2: 0.149 -> 0.184 | Improved |
| add relative positional terms (ablation) | HiST · Brain (NCBI Geo) | R2: 0.131 -> 0.228 | Improved |
| use RoPE instead of default positional terms (ablation) | HiST · Brain (NCBI Geo) | R2: 0.228 -> 0.153 | Hurt |
| use ALiBi instead of default positional terms (ablation) | HiST · Brain (NCBI Geo) | R2: 0.228 -> 0.132 | Hurt |
| add skip connections for multiscale fusion (ablation) | HiST · Brain (NCBI Geo) | R2: 0.197 -> 0.228 | Improved |
| use additive instead of default skip fusion (ablation) | HiST · Brain (NCBI Geo) | R2: 0.228 -> 0.213 | Hurt |
| add downsample-upsample transitions (ablation) | HiST · Brain (NCBI Geo) | R2: 0.213 -> 0.228 | Improved |
| add shifted windows (ablation) | HiST · Brain (NCBI Geo) | R2: 0.217 -> 0.228 | Improved |
| add slide calibration token (ablation) | HiST · Brain (NCBI Geo) | R2: 0.185 -> 0.228 | Improved |
| add Macenko stain normalization on top of calibration token | HiST · Brain (NCBI Geo) | R2: 0.23 -> 0.25 | Improved |
Closest papers by meaning
- Diffusion Generative Modeling for Spatially Resolved Gene Expression Inference from Histology Images (ICLR 2025 · similarity 0.90)
- HEST-1k: A Dataset For Spatial Transcriptomics and Histology Image Analysis (NeurIPS 2024 · similarity 0.89)
- Towards Scalable Context-Aware Single-Cell Spatial Transcriptomics Prediction from Histology Images (NeurIPS 2026 · similarity 0.89)
- Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching (ICML 2025 · similarity 0.89)
- Global Context-aware Representation Learning for Spatially Resolved Transcriptomics (ICML 2025 · similarity 0.87)
- Preserving DEG Rankings for Gene Discovery in Histology-Based Spatial Gene Expression Prediction (NeurIPS 2026 · 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.
