Research map / Architectures and efficiency / RNNs and normalization
Towards Real-Time Neutral Atom Array Assembly via Unsupervised Hologram Generation and Path Optimization
Ge Yan, Yuchen Wang, Junchi Yan
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 FNO bottleneck to U-Net (ablation) | Fourier U-Net · Static-100 | efficiency (%): 7.46 -> 9.40 | Improved |
| add FNO bottleneck to U-Net (ablation) | Fourier U-Net · Static-100 | uniformity (%): 96.19 -> 96.78 | Improved |
| add FNO bottleneck to U-Net (ablation) | Fourier U-Net · Static-400 | efficiency (%): 7.16 -> 8.21 | Improved |
| add FNO bottleneck to U-Net (ablation) | Fourier U-Net · Static-400 | uniformity (%): 85.59 -> 87.48 | Improved |
| add FNO bottleneck to U-Net (ablation) | Fourier U-Net · Static-900 | efficiency (%): 9.76 -> 12.44 | Improved |
| add FNO bottleneck to U-Net (ablation) | Fourier U-Net · Static-900 | uniformity (%): 88.11 -> 89.66 | Improved |
| minimize longest path (MMH) instead of total distance (TDH) | atom-site matching · all five scenarios | maximum moving length: reduced by 18-23% | Improved |
| minimize longest path (MMH) instead of total distance (TDH) | atom-site matching · 60x60 to 45x45 | total survival probability: 1618.76 -> 1684.81 (3% relative) | Improved |
| minimize longest path (MMH) instead of total distance (TDH) | atom-site matching · all scenarios | average minimum distance: marginal decrease (< 1%) | No clear effect |
| use self-supervised Fourier U-Net instead of supervised CNN | hologram generation · S-100 | uniformity (%): 76.43 -> 96.69 | Improved |
| use self-supervised Fourier U-Net instead of supervised CNN | hologram generation · D-900 | uniformity (%): 80.56 -> 81.14 | Improved |
| use Fourier U-Net instead of 30-iteration WGS | hologram generation · S-100 vs D-100 | uniformity (%): S-100 94.53 -> 96.69; D-100 96.50 -> 86.21 | Mixed |
Closest papers by meaning
- Space Group Conditional Flow Matching (NeurIPS 2026 · similarity 0.81)
- Generating Symmetric Materials using Latent Flow Matching (NeurIPS 2026 · similarity 0.81)
- TITAN: A Trajectory-Informed Technique for Adaptive Parameter Freezing in Large-Scale VQE (NeurIPS 2025 · similarity 0.80)
- Unified all-atom molecule generation with neural fields (NeurIPS 2025 · similarity 0.80)
- Scalable Quantum-Inspired Optimization Through Dynamic Qubit Compression (AAAI 2025 · similarity 0.80)
- QuCOOP: A Versatile Framework for Solving Composite and Binary-Parametrised Problems on Quantum Annealers (CVPR 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.
