Research map / Graph learning / Graph algorithms and clustering
Deep Attentive Belief Propagation: Integrating Reasoning and Learning for Solving Constraint Optimization Problems
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 learned dynamic damping factors and neighbor weights rather than static homogeneous damping and equal neighbor weights | DABP versus DBP-SCFG · Random COPs, WGCPs, scale-free networks, and small-world networks | Cost: 27.60 -> 27.12; 30.50 -> 30.02; 32.45 -> 32.04 | Helped |
| Fix neighbor weights uniformly while retaining heterogeneous learned damping factors (ablation) | DABP · COP benchmarks, |X|=60/80/100 | Cost | Hurt |
| Fix neighbor weights uniformly and average damping factors into one homogeneous factor (ablation) | DABP · COP benchmarks, |X|=60/80/100 | Cost | Hurt |
| Use Toulbar2 instead of DABP | DABP · WGCPs, |X|=60 | Cost: 0.18 vs 0.29 | Hurt |
| Train on a larger number of effective iterations (ablation) | DABP · Random COPs and WGCPs, 80 variables | Cost | Hurt |
| Reduce the number of restarts (ablation) | DABP · Random COPs, WGCPs, scale-free networks, and small-world networks | Cost: 27.12 -> 27.16 -> 27.19 | Hurt |
Closest papers by meaning
- Belief Propagation Neural Networks (NeurIPS 2020 · similarity 0.88)
- Finding the Minimal Parameter Budget for Implicit Reasoning: A Data Complexity Driven Scaling Law for Language Models (ICML 2026 · similarity 0.84)
- Learning to Condition: A Neural Heuristic for Scalable MPE Inference (NeurIPS 2025 · similarity 0.84)
- Learning Iterative Reasoning through Energy Diffusion (ICML 2024 · similarity 0.84)
- Are Graph Neural Networks Optimal Approximation Algorithms? (NeurIPS 2024 · similarity 0.84)
- Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous Thought (NeurIPS 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.