Summary of On the Sparsity Of the Strong Lottery Ticket Hypothesis, by Emanuele Natale (coati) et al.
On the Sparsity of the Strong Lottery Ticket Hypothesis
by Emanuele Natale, Davide Ferre’, Giordano Giambartolomei, Frédéric Giroire, Frederik Mallmann-Trenn
First submitted to arxiv on: 18 Oct 2024
Categories
- Main: Machine Learning (stat.ML)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
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 A recent line of research, known as the Strong Lottery Ticket Hypothesis (SLTH), has shown that random neural networks contain subnetworks capable of accurately approximating smaller neural networks without training. This hypothesis was originally motivated by the weaker Lottery Ticket Hypothesis, which states that large random neural networks contain sparse subnetworks that can be trained efficiently to achieve comparable performance. Despite its original motivation, results on the SLTH have not provided guarantees on subnetwork size due to limitations with the Random Subset Sum (RSS) Problem. We provide a proof of the SLTH in classical settings, including dense and equivariant networks, with guarantees on subnetwork sparsity. Our results rely on an essentially tight bound on the Random Fixed-Size Subset Sum Problem (RFSS), a variant of the RSS Problem that asks for subsets of a given size. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A group of researchers studied how big neural networks can be broken down into smaller parts that can do similar tasks without needing to learn. They wanted to know if these smaller parts, called subnetworks, could work together to make predictions as good as the original network. The scientists found that this is possible, but only if the subnetworks are small enough and connected in a special way. |
