Research map / Probabilistic and scientific modeling / Variational and generative sampling
End-to-end Training of Deep Boltzmann Machines by Unbiased Contrastive Divergence with Local Mode Initialization
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 |
|---|---|---|---|
| add unbiased gradient estimation (UCD-LMI) instead of biased PCD-style training (ablation) | DBM without greedy pretraining · MNIST | FID: 112.1 -> 10.33 | Helped |
| add simplified centering trick (map 0 units to -1) (ablation) | DBM with UCD-LMI · MNIST | FID: 30.27 -> 10.33 | Helped |
| add marginalization variance reduction (ablation) | DBM with UCD-LMI · MNIST | FID: 13.70 -> 10.33 | Helped |
| add orthogonal weight initialization (ablation) | DBM with UCD-LMI · MNIST | FID: 18.20 -> 10.33 | Helped |
| replace Gibbs-based coupling with MH-based coupling plus local mode initialization | RBM coupling toy examples · Toy RBM dimensionality 1-200 | coupling time tau + T | Helped |
| initialize Gibbs-based coupling with local mode | Gibbs-based coupling · Toy RBM dimensionality 1-200 | coupling time | No clear effect |
| use DBM training without greedy layer-wise pretraining (UCD-LMI end-to-end) | DBM · MNIST and CIFAR-10 | FID | Helped |
Closest papers by meaning
- BoltzNCE: Learning likelihoods for Boltzmann Generation with Stochastic Interpolants and Noise Contrastive Estimation (NeurIPS 2025 · similarity 0.85)
- Learning with Restricted Boltzmann Machines: Asymptotics of AMP and GD in High Dimensions (NeurIPS 2025 · similarity 0.85)
- Fast training and sampling of Restricted Boltzmann Machines (ICLR 2025 · similarity 0.85)
- Beyond Communication Overhead: A Multilevel Monte Carlo Approach for Mitigating Compression Bias in Distributed Learning (ICML 2025 · similarity 0.85)
- Spherical Boltzmann machines: a solvable theory of learning and generation in energy-based models (NeurIPS 2026 · similarity 0.85)
- Scalable Discrete Diffusion Samplers: Combinatorial Optimization and Statistical Physics (ICLR 2025 · similarity 0.84)
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.