【数据科学学院系列学术活动】“数海领航”大讲堂2026年第13期：From Fourier Embeddings to Spectral Attention: An FSEN Framework for Transformer Architecture Design

来源：

学校官网

收录时间：

2026-07-20 03:01:12

时间：

2026-01-01 00:00:00

地点：

报告人：

学校：

浙江财经大学

关键词：

Transformer, Fourier Embeddings, Spectral Attention, FSEN, Deep Learning, Neural Networks, Sequence Modeling

简介：

This talk introduces the FSEN (Fourier Spectral Embedding Network) framework, a novel approach to Transformer architecture design that leverages Fourier embeddings and spectral attention mechanisms. The framework aims to enhance modeling efficiency and representation capability in sequence learning tasks.

-/- 20