Research map / Optimization / Implicit bias and generalization
Mean Field for the Stochastic Blockmodel: Optimization Landscape and Convergence Issues
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 random i.i.d. initialization of psi instead of an initialization correlated with true communities when p, q are unknown | BCAVI mean field for 2-block SBM · two-class SBM, unknown p,q | convergence to ground truth | Hurt |
| estimate p, q jointly instead of treating them as known | BCAVI mean field for 2-block SBM · two-class SBM | convergence from random initialization | Hurt |
| random initialization with Beta mean near 0.3 or 0.7 instead of mean near 0.16 or 0.71 when p, q known | BCAVI mean field for 2-block SBM · SBM n=200, p=0.4, q=0.025, known p,q | fraction converging to ground truth | Helped |
| initialize with a small mean difference between communities (c0 > 0) | BCAVI mean field for 2-block SBM · SBM n=400, unknown p,q | ||psi - Z||_1 / n | Helped |
Closest papers by meaning
- Principled Parallel Mean-Field Inference for Discrete Random Fields (CVPR 2016 · similarity 0.88)
- Optimal Continuous DR-Submodular Maximization and Applications to Provable Mean Field Inference (ICML 2019 · similarity 0.87)
- Efficient Near-Optimal Testing of Community Changes in Balanced Stochastic Block Models (NeurIPS 2019 · similarity 0.87)
- Multi-Modal Mean-Fields via Cardinality-Based Clamping (CVPR 2017 · similarity 0.87)
- Flexible mean field variational inference using mixtures of non-overlapping exponential families (NeurIPS 2020 · similarity 0.87)
- Online Variational Sequential Monte Carlo (ICML 2024 · 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.