Summary of Wasserstein Wormhole: Scalable Optimal Transport Distance with Transformers, by Doron Haviv et al.
Wasserstein Wormhole: Scalable Optimal Transport Distance with Transformers
by Doron Haviv, Russell Zhang Kunes, Thomas Dougherty, Cassandra Burdziak, Tal Nawy, Anna Gilbert, Dana Pe’er
First submitted to arxiv on: 15 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computational Geometry (cs.CG); Genomics (q-bio.GN)
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 paper introduces Wasserstein Wormhole, a transformer-based autoencoder that embeds empirical distributions into a latent space where Euclidean distances approximate optimal transport (OT) distances. This allows for rapid computation of OT distances, which is particularly useful as cohort size grows. The approach extends multidimensional scaling theory and includes a decoder that maps embeddings back to distributions, enabling operations in the embedding space to generalize to OT spaces. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Wasserstein Wormhole is a new way to compare different groups or types of things. It helps by taking these groups and putting them into a special kind of math problem called an “embedding” that makes it easy to see how similar they are. This tool is useful because it can help us understand and analyze large amounts of data quickly. |
Keywords
» Artificial intelligence » Autoencoder » Decoder » Embedding » Embedding space » Latent space » Transformer
