PBH: A Periodic Binary Head With Two-Stage Training for Efficient Text Classification

Quantized neural networks significantly reduce storage requirements and computational complexity by lowering the numerical precision of weights and activations. Among these, binary neural networks (BNNs) represent an extreme form of quantization, using only 1-bit weights. However, due to the non-dif...

ver descrição completa

Detalhes bibliográficos

Principais autores:	Yanbai Wang, Mengyu Wang, Yingqi Pei, Xueying Jiang, Lu Tan
Formato:	Artigo
Idioma:	Inglês
Publicado em:	IEEE 2026-01-01
coleção:	IEEE Access
Assuntos:	Text classification classification head binary neural networks (BNNs) periodic function two-stage training
Acesso em linha:	https://ieeexplore.ieee.org/document/11435970/
Tags:	Adicionar Tag Sem tags, seja o primeiro a adicionar uma tag!