Research map / Probabilistic and scientific modeling / Variational and generative sampling
Black Box Variational Inference with a Deterministic Objective: Faster, More Accurate, and Even More Black Box
Ryan Giordano, Martin Ingram, Tamara Broderick
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 stochastic ADVI optimization with deterministic SAA objective (DADVI) and second-order optimizer | mean-field ADVI · ARM, Microcredit, Tennis, POTUS | runtime and model evaluations | Improved |
| add linear response covariances on top of DADVI optimum (LRVB) | DADVI · ARM and larger models | posterior standard deviation error | Improved |
| use DADVI instead of mean-field ADVI for posterior means | mean-field ADVI · ARM and larger models | posterior mean error | No clear effect |
| use LRVB instead of full-rank ADVI for covariance | full-rank ADVI · ARM models | posterior standard deviation error | Improved |
| use DADVI instead of RAABBVI for posterior means | RAABBVI · ARM and larger models | posterior mean error | Mixed |
| use stochastic ADVI instead of DADVI for the variational objective | DADVI · ARM and non-ARM models | variational objective at optimum | Hurt |
| lower variational objective value | ADVI methods · ARM and larger models | posterior moment accuracy | No clear effect |
| increase number of draws N in DADVI | DADVI · frequentist coverage | accuracy of sampling variability estimate: N >= 32 | Improved |
| add full-rank ADVI to large models | full-rank ADVI · Occupancy, Tennis, POTUS | feasibility | Hurt |
Closest papers by meaning
- Least squares variational inference (NeurIPS 2025 · similarity 0.90)
- EigenVI: score-based variational inference with orthogonal function expansions (NeurIPS 2024 · similarity 0.90)
- Nearly Dimension-Independent Convergence of Mean-Field Black-Box Variational Inference (NeurIPS 2025 · similarity 0.90)
- Natural Gradient VI: Guarantees for Non-Conjugate Models (NeurIPS 2025 · similarity 0.89)
- Understanding Stochastic Natural Gradient Variational Inference (ICML 2024 · similarity 0.89)
- Provably Scalable Black-Box Variational Inference with Structured Variational Families (ICML 2024 · similarity 0.89)
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.
