Multitask Extension of Geometrically Aligned Transfer Encoder

TL;DR

This paper introduces a multi-task extension of GATE, a geometric encoder, to improve molecular data modeling by leveraging shared information across tasks through geometric alignment.

Contribution

It extends GATE to a multi-task framework, enabling better transfer learning in molecular datasets by aligning geometric representations across tasks.

Findings

Enhanced transfer of information across molecular tasks

Improved performance on molecular datasets

Novel multi-task geometric alignment method

Abstract

Molecular datasets often suffer from a lack of data. It is well-known that gathering data is difficult due to the complexity of experimentation or simulation involved. Here, we leverage mutual information across different tasks in molecular data to address this issue. We extend an algorithm that utilizes the geometric characteristics of the encoding space, known as the Geometrically Aligned Transfer Encoder (GATE), to a multi-task setup. Thus, we connect multiple molecular tasks by aligning the curved coordinates onto locally flat coordinates, ensuring the flow of information from source tasks to support performance on target data.