Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Di-BiLPS: Denoising induced Bidirectional Latent-PDE-Solver under Sparse Observations
Zhonghao Li, Chaoyu Liu, Zhang Qian
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 PDE guidance (ablation) | Di-BiLPS · Darcy Flow, Poisson, bounded NS | relative l2 error: 0.026 -> 0.025 (Helmholtz forward) | No clear effect |
| add PDE guidance (ablation) | Di-BiLPS · Non-bounded NS forward | relative l2 error: 0.044 -> 0.043 | No clear effect |
| add observation guidance (ablation) | Di-BiLPS · Non-bounded NS forward | relative l2 error: 0.274 -> 0.043 | Improved |
| add observation guidance (ablation) | Di-BiLPS · Poisson inverse | relative l2 error: 0.450 -> 0.123 | Improved |
| add condition guidance (ablation) | Di-BiLPS · Poisson forward | relative l2 error: 1.185 -> 0.038 | Improved |
| add condition guidance (ablation) | Di-BiLPS · Non-bounded NS inverse | relative l2 error: 1.429 -> 0.116 | Improved |
| raise observation rate from 1% to 3% to 10% | Di-BiLPS · Darcy Flow forward | relative l2 error: 0.025 -> 0.021 -> 0.020 | Improved |
| raise observation rate from 1% to 3% to 10% | Di-BiLPS · Poisson inverse | relative l2 error: 0.145 -> 0.123 -> 0.117 | Improved |
| use Di-BiLPS instead of DiffusionPDE | Di-BiLPS · Non-bounded NS inverse | relative l2 error: 0.104 -> 0.116 | Hurt |
| use Di-BiLPS instead of DiffusionPDE | Di-BiLPS · Helmholtz forward | relative l2 error: 0.088 -> 0.025 | Improved |
| use Di-BiLPS instead of DiffusionPDE | Di-BiLPS · Darcy Flow forward | runtime (s): 255.74 -> 35.45 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DiffusionPDE — DiffusionPDE: Generative PDE-Solving under Partial Observation (NeurIPS 2024) · better on 8 of 10 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
- CALM-PDE: Continuous and Adaptive Convolutions for Latent Space Modeling of Time-dependent PDEs (NeurIPS 2025 · similarity 0.89)
- Guided Diffusion Sampling on Function Spaces with Applications to PDEs (NeurIPS 2025 · similarity 0.87)
- DPOT: Auto-Regressive Denoising Operator Transformer for Large-Scale PDE Pre-Training (ICML 2024 · similarity 0.87)
- Noise Conditional Variational Score Distillation (ICML 2025 · similarity 0.87)
- DiffusionPDE: Generative PDE-Solving under Partial Observation (NeurIPS 2024 · similarity 0.86)
- SBNO : Schrödinger Bridge Neural Operator for the Forward and Inverse Problems under Missing Information (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.
