Research map / Probabilistic and scientific modeling / Variational and generative sampling
On Solving Probabilistic Linear Diophantine Equations
Patrick Kreitzberg, Oliver Serang
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 trimming to p-convolution trees (ablation) | p-convolution tree · Y=X1+...+Xm binary, m=32768 | runtime (s): 890.6 -> 1.421 | Improved |
| add trimming to p-convolution trees (ablation) | p-convolution tree · Y=X1+...+Xm binary, m=2 | runtime (s): 0.0001660 -> 0.0001700 | No clear effect |
| add trimming to p-convolution trees (ablation) | p-convolution tree · HOP-MAP comparison, m=16384, p=inf | runtime (s): 3218.0 -> 1.372 | Improved |
| add lazy trimming to p-convolution trees (ablation) | Fido protein inference · 18mix, p=1 | runtime (s): 12.48 -> 2.94 | Improved |
| add lazy trimming to p-convolution trees (ablation) | Fido protein inference · Yeast, p=1 | runtime (s): 43.35 -> 40.82 | Improved |
| add lazy trimming to p-convolution trees (ablation) | Fido protein inference · IPRG, p=inf | runtime (s): 23.31 -> 12.23 | Improved |
| use trimmed tree instead of specialized HOP-MAP (ablation) | max-marginal on Y · m=16384, p=inf | runtime (s): 0.05271 -> 1.372 | Hurt |
| use trimmed tree instead of specialized noisy-or (ablation) | noisy-or on Y · m=16384, p=1 | runtime (s): 0.05118 -> 0.5097 | Hurt |
| hold out Xi from the Cartesian product (ablation) | CPT · m=256 n=32, k=4096 | runtime (s): 0.04264 -> 0.004865 | Improved |
| hold out Xi and compute only the marginal on Xi (ablation) | CPT · m=256 n=32, k=4096 | runtime (s): 0.004865 -> 0.0007699 | Improved |
| hold out Xi from the Cartesian product (ablation) | CPT · m=128 n=32, k=4096, Y | MSE: 1.675 × 10−05 -> 6.073 × 10−06 | Improved |
| use CPT holdout Xi instead of FFT Evergreen (ablation) | max-marginal · m=256 n=32, k=4096 | runtime (s): 0.5504 -> 0.004865 | Improved |
Closest papers by meaning
- d$^2$p: Structured Soft Attention Is All You Need (ICML 2026 · similarity 0.85)
- Exact Sampling with Integer Linear Programs and Random Perturbations (AAAI 2016 · similarity 0.85)
- Maximum A Posteriori Inference in Sum-Product Networks (AAAI 2018 · similarity 0.85)
- On the Expressive Power of Tree-Structured Probabilistic Circuits (NeurIPS 2024 · similarity 0.84)
- Neural Network Approximators for Marginal MAP in Probabilistic Circuits (AAAI 2024 · similarity 0.83)
- Composing Biases by Using CP to Decompose Minimal Functional Dependencies for Acquiring Complex Formulae (AAAI 2024 · similarity 0.83)
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.
