Research map / Probabilistic and scientific modeling / Variational and generative sampling
Challenges and Opportunities in High Dimensional Variational 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 |
|---|---|---|---|
| apply PSIS correction to exclusive KL approximation (ablation) | BBVI exclusive KL · posteriordb radon | predictive log-likelihood: -353.0 -> -325.0 | Helped |
| apply PSIS correction to exclusive KL approximation (ablation) | BBVI exclusive KL · posteriordb diamonds | predictive log-likelihood: -2.6 -> 1.5 | Helped |
| apply PSIS correction to exclusive KL approximation (ablation) | BBVI exclusive KL · posteriordb mesquite | predictive log-likelihood: -2512 -> -5418 | Hurt |
| apply PSIS correction to exclusive KL approximation (ablation) | BBVI exclusive KL · posteriordb nes1996 | predictive log-likelihood: -412.8 -> -427.9 | Hurt |
| apply PSIS correction to inclusive KL approximation (ablation) | BBVI inclusive KL · posteriordb dogs | predictive log-likelihood: -110 -> -70.5 | Helped |
| apply PSIS correction to inclusive KL approximation (ablation) | BBVI inclusive KL · posteriordb nes1996 | predictive log-likelihood: -2140.5 -> -499.3 | Helped |
| use exclusive KL instead of inclusive KL | BBVI · posteriordb diamonds | predictive log-likelihood: -3196.6 -> -2.6 | Helped |
| use exclusive KL instead of inclusive KL | BBVI · posteriordb radon | predictive log-likelihood: -377.4 -> -353.0 | Helped |
| use exclusive KL instead of inclusive KL | BBVI · posteriordb dogs | predictive log-likelihood: -110 -> -71.2 | Helped |
| use exclusive KL instead of inclusive KL in low dimension | BBVI · posteriordb arK (D=7) | predictive log-likelihood: -34.3 -> -35.2 | Hurt |
| increase target dimension D from 2 to 50 with inclusive KL | mean-field and flow BBVI · robust regression | Pareto k-hat | Hurt |
| use normalizing-flow family with inclusive KL | BBVI · robust regression | largest converged dimension: D = 2 | Hurt |
Closest papers by meaning
- Provably Scalable Black-Box Variational Inference with Structured Variational Families (ICML 2024 · similarity 0.92)
- Nearly Dimension-Independent Convergence of Mean-Field Black-Box Variational Inference (NeurIPS 2025 · similarity 0.92)
- EigenVI: score-based variational inference with orthogonal function expansions (NeurIPS 2024 · similarity 0.91)
- Efficient Mixture Learning in Black-Box Variational Inference (ICML 2024 · similarity 0.91)
- Batch and match: black-box variational inference with a score-based divergence (ICML 2024 · similarity 0.90)
- Least squares variational inference (NeurIPS 2025 · similarity 0.90)
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.