Research map / Architectures and efficiency / RNNs and normalization
Explaining Grokking and Information Bottleneck through Neural Collapse Emergence
Keitaro Sakamoto, Issei Sato
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| train longer past training-loss convergence so within-class variance (RNC1) contracts | 4-layer MLP on MNIST with 8x init scale · MNIST grokking | test accuracy (%): 15 -> 88 | Improved |
| increase weight decay lambda (ablation) | MLP on MNIST · MNIST grokking | timing of grokking and RNC1 decrease | Improved |
| increase weight decay lambda (ablation) | MLP on MNIST · MNIST IB dynamics | redundant information (MI, nHSIC) | Improved |
Closest papers by meaning
- Flatness is Necessary, Neural Collapse is Not: Rethinking Generalization via Grokking (NeurIPS 2025 · similarity 0.90)
- Understanding and Leveraging the Learning Phases of Neural Networks (AAAI 2024 · similarity 0.89)
- Grokking as a First Order Phase Transition in Two Layer Networks (ICLR 2024 · similarity 0.88)
- Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks (ICML 2025 · similarity 0.88)
- Grokking as the transition from lazy to rich training dynamics (ICLR 2024 · similarity 0.88)
- A Dynamical Model of Neural Scaling Laws (ICML 2024 · similarity 0.88)
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.
