Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Matrix Product States for Inference in Discrete Probabilistic Models
Rasmus Bonnevie, Mikkel N. Schmidt
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 |
|---|---|---|---|
| use unbiased RELAX gradients with a learned control variate instead of the biased differentiable MPS (Gumbel-softmax) (ablation) | MPS variational posterior, stochastic block model · Karate network subgraph (N=9, K=2) | true ELBO | Improved |
| use the Gumbel-softmax temperature 0.5 instead of 0.1 in direct dMPS gradients (ablation) | dMPS gradient estimator · N=4 cycle graph, K=2 | gradient variance and bias | Mixed |
| add RELAX variance reduction with a learned MPS control variate over the plain score estimator (ablation) | score gradient estimator · N=4 cycle graph, K=2 | gradient variance | Improved |
| increase MPS rank (ablation) | MPS approximation · Erdos-Renyi graphs with 4, 6, 8 vertices | KL divergence to true posterior | Improved |
| increase MPS rank (ablation) | MPS sampling · order-24 binary MPS | time to draw one sample | Hurt |
Closest papers by meaning
- Variational Inference in Mixed Probabilistic Submodular Models (NeurIPS 2016 · similarity 0.88)
- Probabilistic Circuits for Variational Inference in Discrete Graphical Models (NeurIPS 2020 · similarity 0.87)
- Universal Boosting Variational Inference (NeurIPS 2019 · similarity 0.87)
- Continuous Mixtures of Tractable Probabilistic Models (AAAI 2023 · similarity 0.87)
- Expressive power of tensor-network factorizations for probabilistic modeling (NeurIPS 2019 · similarity 0.87)
- Variational Inference with Mixtures of Isotropic Gaussians (NeurIPS 2025 · similarity 0.86)
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.
