Research map / Learning theory and algorithms / Sample complexity
Exponential Reduction in Sample Complexity with Learning of Ising Model Dynamics
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 one-step Glauber samples from the M-regime rather than a quickly mixing trajectory in the T-regime | D-RISE and D-RPLE · Ising models; studied lattice and random regular graph topologies | sample-complexity β scaling: 80% and 90% | Helped |
| Use entropy-based active selection of initial configurations | D-RISE · Ferromagnetic periodic lattice with weak antiferromagnetic impurity | number of samples required for successful graph reconstruction: 47% | Helped |
| Use entropy-based active selection of initial configurations | D-RISE · Ferromagnetic periodic lattice with weak antiferromagnetic impurity | sample-complexity scaling with β | No clear effect |
| Learn from T-regime dynamics rather than i.i.d. samples | D-RISE · Spin glass on a periodic lattice | sample-complexity β scaling: 35% | Helped |
| Learn from T-regime dynamics rather than i.i.d. samples | D-RISE · Spin glass on a random regular graph | sample-complexity β scaling: 27% | Helped |
| Learn from T-regime dynamics rather than i.i.d. samples | D-RISE and D-RPLE · Ferromagnetic models on lattices and random regular graphs | sample-complexity β scaling | No clear effect |
| Use M-regime dynamics on random regular graphs | D-RISE and D-RPLE · Random regular graphs | sample-complexity dependence on β: a constant number of samples independent of β | Helped |
Closest papers by meaning
- Interaction Screening: Efficient and Sample-Optimal Learning of Ising Models (NeurIPS 2016 · similarity 0.88)
- Efficient Learning of Discrete Graphical Models (NeurIPS 2020 · similarity 0.87)
- Cross-fluctuation phase transitions reveal sampling dynamics in diffusion models (NeurIPS 2025 · similarity 0.85)
- Fast Wasserstein rates for estimating probability distributions of probabilistic graphical models (NeurIPS 2026 · similarity 0.85)
- Dynamical Decoupling of Generalization and Overfitting in Large Two-Layer Networks (NeurIPS 2025 · similarity 0.85)
- Wasserstein Gradient Flows and Forward-Only Diffusion Are Not Enough for Multimodal Sampling (NeurIPS 2026 · 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.