Research map / Probabilistic and scientific modeling / Variational and generative sampling
Variational methods for simulation-based inference
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use SNVI instead of SNLE with 100-chain MCMC | Pyloric network inference · Pyloric network | runtime for 10k samples: 808 minutes -> 11 minutes | Helped |
| Use SNVI+fKL rather than previous methods | Pyloric network inference · Pyloric network | simulations to obtain posterior: 350,000; 25 times fewer than previous methods | Helped |
| Use posterior samples rather than prior samples | Pyloric network simulator · Pyloric network | valid simulation rate: 1% -> 86% | Helped |
| Use SNVI+fKL instead of 100-chain MCMC | Two moons · Two moons | runtime: up to twenty times faster | Helped |
| Use mass-covering SNVI objectives instead of reverse-KL SNPLA | Two moons and SLCP inference · Two moons and SLCP | C2ST | Helped |
| Add SIR posterior refinement (ablation) | All SNVI variants · Bernoulli GLM, Lotka Volterra, Two moons, and SLCP | C2ST | Helped |
| Use reverse KL rather than a mass-covering objective (ablation) | SNVI / SNPLA · Two moons and SLCP | posterior approximation quality | Hurt |
| Choose fKL, IW-ELBO, or α-divergence as the mass-covering objective (ablation) | SNVI · Bernoulli GLM, Lotka Volterra, Two moons, and SLCP | C2ST after SIR | No clear effect |
| Use likelihood estimation with SNVI instead of SNLE with MCMC | SBI benchmark tasks · Bernoulli GLM, Lotka Volterra, Two moons, and SLCP | runtime | Helped |
Closest papers by meaning
- Brain-like Variational Inference (NeurIPS 2025 · similarity 0.89)
- Closing the Approximation Gap in Simulation-free Latent SDEs (NeurIPS 2026 · similarity 0.88)
- VIKING: Deep variational inference with stochastic projections (NeurIPS 2025 · similarity 0.88)
- Variational Inference with Coverage Guarantees in Simulation-Based Inference (ICML 2024 · similarity 0.87)
- Globally Convergent Variational Inference (NeurIPS 2024 · similarity 0.87)
- Effortless, Simulation-Efficient Bayesian Inference using Tabular Foundation Models (NeurIPS 2025 · similarity 0.87)
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.