Summary of Crystalformer: Infinitely Connected Attention For Periodic Structure Encoding, by Tatsunori Taniai et al.
Crystalformer: Infinitely Connected Attention for Periodic Structure Encoding
by Tatsunori Taniai, Ryo Igarashi, Yuta Suzuki, Naoya Chiba, Kotaro Saito, Yoshitaka Ushiku, Kanta Ono
First submitted to arxiv on: 18 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Materials Science (cond-mat.mtrl-sci); Computational Physics (physics.comp-ph)
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 proposes a novel approach for predicting physical properties of materials from their crystal structures using a Transformer-based encoder architecture called Crystalformer. The approach, which is based on neural potential summation, leverages the infinitely repeating structure of crystals to develop a computationally tractable formulation. Compared to an existing Transformer-based model, Crystalformer requires fewer parameters and outperforms state-of-the-art methods for various property regression tasks on two benchmark datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper tries to make predictions about materials based on their crystal structures. It’s like trying to guess what a material will be good for based on its molecular structure, but instead of molecules, it’s using the repeating patterns found in crystals. The method is called Crystalformer and uses something called neural potential summation to do this. It’s simpler than other methods that already exist and actually does better at predicting things. |
Keywords
* Artificial intelligence * Encoder * Regression * Transformer
