ToaSt: Token Channel Selection and Structured Pruning for Efficient ViT 文章

ArXiv CS.CV2026-06-16NEWSen作者: Hyunchan Moon, Cheonjun Park, Steven L. Waslander

大语言模型

查看原文 →

ToaSt: Token Channel Selection and Structured Pruning for Efficient ViT · 相关技术

相关技术

attention operation coupled head-wise structured pruning token compression structured weight pruning Vision Transformers (ViTs)Token Channel Selection ToaST Feed-Forward Network Multi-Head Self-Attention