Research map / Probabilistic and scientific modeling / Variational and generative sampling
All in the Exponential Family: Bregman Duality in Thermodynamic 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 |
|---|---|---|---|
| Use adaptive moment-spacing instead of grid search to choose the intermediate distribution | TVO · Omniglot | test log likelihood | No clear effect |
| Use adaptive moment-spacing instead of grid search to choose the intermediate distribution | TVO · MNIST | test log likelihood | No clear effect |
| Use a properly chosen intermediate distribution rather than the ELBO alone | TVO · Omniglot | test log likelihood | Helped |
| Use a properly chosen intermediate distribution rather than the ELBO alone | TVO · MNIST | test log likelihood | Helped |
| Vary the number of intermediate distributions under moment-spacing (ablation) | TVO · Omniglot | test log likelihood | Hurt |
| Use moment-spacing rather than linear or log-uniform spacing (ablation) | TVO · Omniglot | test log likelihood | Mixed |
| Use doubly-reparameterized gradients instead of REINFORCE-style gradients (ablation) | TVO · Omniglot | test log likelihood | Helped |
| Use doubly-reparameterized gradients instead of REINFORCE-style gradients (ablation) | TVO · Omniglot | posterior KL divergence | Helped |
| Use doubly-reparameterized gradients with moment-spacing rather than IWAE | TVO · MNIST | test log likelihood | No clear effect |
| Use doubly-reparameterized gradients with moment-spacing rather than IWAE DREG | TVO · MNIST | posterior inference | Helped |
| Use equal spacing in KL or symmetrized KL instead of the proposed moment-spacing schedule (ablation) | TVO · Omniglot | model learning performance | No clear effect |
Closest papers by meaning
- Asymptotics of Alpha-Divergence Variational Inference Algorithms with Exponential Families (NeurIPS 2024 · similarity 0.87)
- Particle Semi-Implicit Variational Inference (NeurIPS 2024 · similarity 0.87)
- Efficient Mixture Learning in Black-Box Variational Inference (ICML 2024 · similarity 0.87)
- Theoretical Guarantees for Variational Inference with Fixed-Variance Mixture of Gaussians (ICML 2024 · similarity 0.86)
- Generalized Variational Inference via Optimal Transport (AAAI 2024 · similarity 0.85)
- Model-Informed Flows for Bayesian Inference (NeurIPS 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.