Research map / Probabilistic and scientific modeling / Variational and generative sampling
Learning Weighted Model Integration Distributions
Paolo Morettin, Samuel Kolb, Stefano Teso, Andrea Passerini
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 LARIAT (learned support plus normalization) to DET | DET · UCI/MLC++ anneal-U | test log-likelihood: −61.2 -> −38.5 | Improved |
| Add LARIAT to MSPN | MSPN · UCI/MLC++ anneal-U | test log-likelihood: −41.9 -> 6.0 | Improved |
| Add LARIAT to DET | DET · UCI/MLC++ solar | test log-likelihood: −15.0 -> −2.6 | Improved |
| Add LARIAT to MSPN | MSPN · UCI/MLC++ balance-scale | test log-likelihood: −7.5 -> −6.4 | No clear effect |
| Add LARIAT to DET | DET · UCI/MLC++ breast | test log-likelihood: −29.6 -> −29.2 | No clear effect |
| Add LARIAT across all 18 hybrid UCI datasets | DET and MSPN · UCI/MLC++ | test log-likelihood | Improved |
| INCAL+ learned support instead of trivial range support | INCAL+ · synthetic datasets | misclassification error | Improved |
| INCAL+ with timeout on harder instances | INCAL+ · synthetic datasets, l = 4 | misclassification error | Hurt |
| Add LARIAT with ground-truth support | DET and MSPN · synthetic datasets | IAE | Improved |
| Add LARIAT with INCAL+ learned support | DET · synthetic datasets | IAE | Improved |
| Add LARIAT with INCAL+ learned support | MSPN · synthetic datasets | IAE | No clear effect |
| Local normalization timeout of 1200s then fallback to global normalization | LARIAT-DET · complex UCI tasks | test log-likelihood | No clear effect |
Closest papers by meaning
- LT-Soups: Bridging Head and Tail Classes via Subsampled Model Soups (NeurIPS 2025 · similarity 0.82)
- Simultaneous identification of models and parameters of scientific simulators (ICML 2024 · similarity 0.82)
- Stationary Latent Weight Inference for Unreliable Observations from Online Test-Time Adaptation (ICML 2024 · similarity 0.81)
- Partition Tree: Conditional Density Estimation over General Outcome Spaces (NeurIPS 2026 · similarity 0.81)
- Total Variation Distance Meets Probabilistic Inference (ICML 2024 · similarity 0.81)
- Scaling laws for learning with real and surrogate data (NeurIPS 2024 · similarity 0.81)
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.
