Research map / Probabilistic and scientific modeling / Molecules and proteins
NAOMI: Non-Autoregressive Multiresolution Sequence Imputation
Yukai Liu, Rose Yu, Stephan Zheng, Eric Zhan, Yisong Yue
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 |
|---|---|---|---|
| decode non-autoregressively at multiple resolutions instead of single-resolution autoregressive decoding (ablation) | NAOMI vs SingleRes · PEMS-SF traffic | L2 loss (1e-4): 4.51 -> 3.54 | Improved |
| decode non-autoregressively at multiple resolutions instead of single-resolution autoregressive decoding (ablation) | NAOMI vs SingleRes · Billiards | L2 loss (1e-2): 0.233 -> 0.067 | Improved |
| decode non-autoregressively at multiple resolutions instead of single-resolution autoregressive decoding (ablation) | NAOMI vs SingleRes · Billiards | step change (1e-3): 9.290 -> 7.239 | Improved |
| decode non-autoregressively at multiple resolutions instead of single-resolution autoregressive decoding (ablation) | NAOMI vs SingleRes · Basketball | Step Change (1e-3): 4.811 -> 2.565 | Improved |
| decode non-autoregressively at multiple resolutions instead of single-resolution autoregressive decoding (ablation) | NAOMI vs SingleRes · Basketball | Player Distance: 0.417 -> 0.423 | No clear effect |
| use multiresolution NAOMI instead of linear interpolation | NAOMI · Billiards | step change (1e-3): 0.961 -> 7.239 | Hurt |
| use NAOMI instead of KNN | NAOMI · Basketball | OOB Rate (1e-3): 0.128 -> 1.733 | Hurt |
| increase proportion of missing values (ablation) | NAOMI and SingleRes · Billiards | imputation quality | Hurt |
| apply per-step discriminator with adversarial imitation (MaskGAN) vs whole-sequence discriminator (GRUI) | imputation baselines · PEMS-SF traffic | L2 loss (1e-4): 15.24 -> 6.02 | Improved |
Later papers that compared with this method
1 later accepted paper reports NAOMI as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Imitative Non-Autoregressive Modeling for Trajectory Forecasting and Imputation (CVPR 2020 · better on 3 of 3 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Zero-shot Imputation with Foundation Inference Models for Dynamical Systems (ICLR 2025 · similarity 0.87)
- Diffusion Transformers for Imputation: Statistical Efficiency and Uncertainty Quantification (NeurIPS 2025 · similarity 0.86)
- ProCTI: Prototype-Refined Global Conditioning for Diffusion-Based Time Series Imputation (NeurIPS 2026 · similarity 0.85)
- RDPI: A Refine Diffusion Probability Generation Method for Spatiotemporal Data Imputation (AAAI 2025 · similarity 0.85)
- Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction (NeurIPS 2024 · similarity 0.84)
- Discretizing Continuous Time Series for Imputation with Masked Diffusion Training (NeurIPS 2026 · similarity 0.84)
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.
