Summary of The Prevalence Of Neural Collapse in Neural Multivariate Regression, by George Andriopoulos et al.
The Prevalence of Neural Collapse in Neural Multivariate Regression
by George Andriopoulos, Zixuan Dong, Li Guo, Zifan Zhao, Keith Ross
First submitted to arxiv on: 6 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The abstract discusses Neural Collapse (NC) and its extension, Neural Regression Collapse (NRC), which occurs during the final stage of training for classification and regression problems. Specifically, multivariate regression exhibits NRC1-3: collapsing last-layer feature vectors to the subspace spanned by principal components, weight vectors, and a functional form dependent on target covariance. The authors empirically demonstrate this phenomenon across various datasets and network architectures. They then provide an explanation by modeling the regression task using the Unconstrained Feature Model (UFM), showing that NRC1-3 emerges when regularization parameters are strictly positive. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Neural networks can get stuck in a certain way during training, called Neural Collapse. Researchers found that this happens not just for classification problems, but also for regression problems, where the goal is to predict continuous values. They named this phenomenon Neural Regression Collapse (NRC). By studying NRC, scientists hope to understand how neural networks work and why they sometimes get stuck. |
Keywords
» Artificial intelligence » Classification » Regression » Regularization
