Research map / Optimization / Implicit bias and generalization
Deterministic PAC-Bayesian generalization bounds for deep networks via generalizing noise-resilience
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 the proposed noise-resilience framework instead of product-of-spectral-norm-based bounds | ReLU feedforward networks · MNIST, networks with H=40 | bound depth-scaling rate: 2.15^D -> 1.57^D | Helped |
| Use the proposed bound rather than existing bounds | ReLU feedforward networks · D=28 | generalization-bound value | Helped |
| Use the proposed bound rather than existing bounds | ReLU feedforward networks · Networks of varying depth | generalization-bound value | Hurt |
Closest papers by meaning
- PAC-Bayes Compression Bounds So Tight That They Can Explain Generalization (NeurIPS 2022 · similarity 0.91)
- A PAC-Bayesian Approach to Generalization Bounds for Graph Neural Networks (ICLR 2021 · similarity 0.90)
- Generalization Bounds via Meta-Learned Model Representations: PAC-Bayes and Sample Compression Hypernetworks (ICML 2025 · similarity 0.90)
- On the Importance of Gradient Norm in PAC-Bayesian Bounds (NeurIPS 2022 · similarity 0.90)
- Non-vacuous Generalization Bounds at the ImageNet Scale: a PAC-Bayesian Compression Approach (ICLR 2019 · similarity 0.90)
- Generalizability of Neural Networks Minimizing Empirical Risk Based on Expressive Power (ICLR 2025 · similarity 0.89)
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.