Summary of Emp: Enhance Memory in Data Pruning, by Jinying Xiao and Ping Li and Jie Nie and Zhe Tang
EMP: Enhance Memory in Data Pruning
by Jinying Xiao, Ping Li, Jie Nie, Zhe Tang
First submitted to arxiv on: 28 Aug 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 This paper addresses the issue of low-frequency learning (LFL) in large language and vision models, which occurs when dataset pruning is used to reduce training costs. LFL causes the model to forget most samples as the pruning rate increases. To address this, the authors decompose the scoring function of LFL, provide a theoretical explanation for its inefficiency, and propose adding a memory term to enhance the model’s memory capability. They also explore memory in self-supervised learning (SSL) and derive a memory term both theoretically and experimentally. The proposed Enhance Memory Pruning (EMP) method addresses the issue of insufficient memory under high pruning rates by enhancing the model’s memory of data, improving its performance. The authors evaluate EMP on tasks such as image classification, natural language understanding, and model pre-training, showing that it outperforms current methods with 2.2% improvement in CIFAR100-ResNet50 pre-training task at 70% pruning rate. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary LFL prevents large language and vision models from remembering most samples when dataset pruning is used to reduce training costs. The authors propose adding a memory term to the scoring function to enhance the model’s memory capability, as well as exploring memory in self-supervised learning (SSL). They also develop Enhance Memory Pruning (EMP), which improves model performance under extreme pruning rates. |
Keywords
» Artificial intelligence » Image classification » Language understanding » Pruning » Self supervised
