Many Property
Research on "many property" problems focuses on predicting or explaining multiple properties simultaneously, moving beyond single-property analyses. Current efforts concentrate on developing and improving multimodal deep learning models, such as transformer-based architectures and diffusion models, along with refining explanation methods like Shapley values and investigating the properties of various kernel-based approaches. This research is significant because it addresses the limitations of single-property models and enables more comprehensive understanding and prediction in diverse fields, including materials science, drug discovery, and climate modeling.
Papers
ReMax: A Simple, Effective, and Efficient Reinforcement Learning Method for Aligning Large Language Models
Ziniu Li, Tian Xu, Yushun Zhang, Zhihang Lin, Yang Yu, Ruoyu Sun, Zhi-Quan Luo
On the Properties and Estimation of Pointwise Mutual Information Profiles
Paweł Czyż, Frederic Grabowski, Julia E. Vogt, Niko Beerenwinkel, Alexander Marx
From axioms over graphs to vectors, and back again: evaluating the properties of graph-based ontology embeddings
Fernando Zhapa-Camacho, Robert Hoehndorf
A Comprehensive and Versatile Multimodal Deep Learning Approach for Predicting Diverse Properties of Advanced Materials
Shun Muroga, Yasuaki Miki, Kenji Hata