Research map / Graph learning / GNN architectures
Transition to Linearity of General Neural Networks with Directed Acyclic Graph Architecture
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 |
|---|---|---|---|
| Increase the minimum in-degree m while keeping maximum in-degree polynomially bounded | Feedforward DAG networks · General feedforward DAG networks | Hessian spectral norm: \tilde{O}(1/\sqrt{m}) | Helped |
| Remove edges independently with probability 1/2 | 3-layer DenseNet · CIFAR-2 (10 data points) | maximum relative NTK change during training: \Theta(1/\sqrt{m}) | Helped |
| Add a nonlinear activation to the output neurons | Wide feedforward networks · General feedforward networks | Hessian spectral norm: O(1) | Hurt |
| Allow two bottleneck neurons with in-degree 1 while the number of neurons grows | Feedforward DAG networks · Networks with bottleneck neurons | transition to linearity | No clear effect |
Closest papers by meaning
- On Vanishing Gradients, Over-Smoothing, and Over-Squashing in GNNs: Bridging Recurrent and Graph Learning (NeurIPS 2025 · similarity 0.86)
- What do Graph Neural Networks learn? Insights from Tropical Geometry (NeurIPS 2024 · similarity 0.84)
- How Graph Neural Networks Learn: Lessons from Training Dynamics (ICML 2024 · similarity 0.84)
- Graph neural networks and non-commuting operators (NeurIPS 2024 · similarity 0.84)
- Flat Channels to Infinity in Neural Loss Landscapes (NeurIPS 2025 · similarity 0.84)
- Universal Sharpness Dynamics in Neural Network Training: Fixed Point Analysis, Edge of Stability, and Route to Chaos (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.