Research map / Architectures and efficiency / Attention and transformers
Attend and Predict: Understanding Gene Regulation by Selective Attention on Chromatin
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 |
|---|---|---|---|
| Use the hierarchical LSTM-α, β model instead of DeepChrome | AttentiveChrome · REMC, 56 cell types | AUC: 36 | Helped |
| Remove HM-level LSTM and β-attention, using concatenation and an MLP instead (ablation) | AttentiveChrome · REMC, 56 cell types | mean AUC: 0.8115 -> 0.8133 | No clear effect |
| Use CNN encoders instead of LSTMs in the hierarchical attention model (ablation) | AttentiveChrome · REMC, 56 cell types | mean AUC: 0.8115 -> 0.7936 | Hurt |
| Use one attention layer over CNN input and remove max-pooling (ablation) | DeepChrome (CNN) · REMC, 56 cell types | mean AUC: 0.8008 -> 0.7622 | Hurt |
| Add attention to the baseline LSTM (ablation) | LSTM · REMC, 56 cell types | mean AUC: 0.8052 -> 0.8100 | Helped |
| Use hierarchical attention maps rather than class-based optimization for interpretation | AttentiveChrome · H1-hESC | Pearson correlation with Hactive: 0.0562 -> 0.8995 | Helped |
| Use hierarchical attention maps rather than saliency maps for interpretation | AttentiveChrome · H1-hESC | Pearson correlation with Hactive: 0.1822 -> 0.8995 | Helped |
Closest papers by meaning
- Learning to Discover Regulatory Elements for Gene Expression Prediction (ICLR 2025 · similarity 0.83)
- ViTally Consistent: Scaling Biological Representation Learning for Cell Microscopy (ICML 2025 · similarity 0.82)
- STRAND: Sequence-Conditioned Transport for Single-Cell Perturbations (NeurIPS 2026 · similarity 0.82)
- KINDLE: Knowledge-Guided Distillation for Prior-Free Gene Regulatory Network Inference (NeurIPS 2025 · similarity 0.82)
- ChromFound: Towards A Universal Foundation Model for Single-Cell Chromatin Accessibiltiy Data (NeurIPS 2025 · similarity 0.82)
- scMBERT: A Pre-Trained Deep Learning Model for Single-Cell Multiomic Data Representation and Prediction (Student Abstract) (AAAI 2025 · similarity 0.82)
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.