Research map / Probabilistic and scientific modeling / Molecules and proteins
Scalable Simulation-Based Model Inference with Test-Time Complexity Control
Manuel Gloeckler, J.P. Manzano-Patrón, Stamatios Sotiropoulos, Cornelius Schröder, Jakob H. Macke
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 |
|---|---|---|---|
| replace SBI_joint (restricted model choice) with PRISM for tractography | dMRI fiber orientation posteriors · dMRI tractography vs MCMC | mean tract-wise correlation with MCMC: 0.86 -> 0.96 | Improved |
| increase number of components K (larger model space) | PRISM · Symbolic regression, 200-model subspace | top-1 accuracy: 0.796 -> 0.503 | Hurt |
| increase number of components K (larger model space) | PRISM · Symbolic regression, 200-model subspace | top-5 accuracy: 0.940 at K=80 and 0.905 at K=100 | Hurt |
| increase network capacity (model_dim 32, 64, 128) (ablation) | PRISM · Symbolic regression, large model spaces | rKSD / rRMSE | Improved |
| grow model space beyond 2^30 with fixed 24h training | PRISM · Symbolic regression | rRMSE | Hurt |
| restrict evaluation to a single fixed (overly simple) model instead of model selection (ablation) | PRISM on dMRI B3S · dMRI UKB | rKSD | Hurt |
| use PRISM compared with SBMI at K=15 | symbolic regression, fixed model prior · Symbolic regression K=15 | model/parameter inference quality | Improved |
| use Bayesian model average of inferred models instead of DTI or Rumba | dMRI extended model family (BSZT) · dMRI UKB | reconstruction RMSE and LOOCV | Improved |
Closest papers by meaning
- Simultaneous identification of models and parameters of scientific simulators (ICML 2024 · similarity 0.88)
- All-in-one simulation-based inference (ICML 2024 · similarity 0.87)
- Effortless, Simulation-Efficient Bayesian Inference using Tabular Foundation Models (NeurIPS 2025 · similarity 0.86)
- Encoder-Decoder Diffusion Language Models for Efficient Training and Inference (NeurIPS 2025 · similarity 0.85)
- Derivative-Free Guidance in Continuous and Discrete Diffusion Models with Soft Value-based Decoding (NeurIPS 2025 · similarity 0.85)
- Compositional simulation-based inference for time series (ICLR 2025 · similarity 0.85)
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.
