paper-with-me

홈 › Papers

Multi-Head Attention with Disagreement Regularization

2018-10-24 · EMNLP 2018 10 · Jian Li, Zhaopeng Tu, Baosong Yang, Michael R. Lyu, Tong Zhang

Multi-head attention is appealing for the ability to jointly attend to information from different representation subspaces at different positions. In this work, we introduce a disagreement regularization to explicitly encourage the diversity among multiple attention heads. Specifically, we propose three types of disagreement regularization, which respectively encourage the subspace, the attended positions, and the output representation associated with each attention head to be different from other heads. Experimental results on widely-used WMT14 English-German and WMT17 Chinese-English translation tasks demonstrate the effectiveness and universality of the proposed approach.

📄 PDF Abstract BibTeX arXiv:1810.10183

Code (0)

등록된 구현이 없습니다.

Tasks

DiversityTranslation

Similar Papers 제목 키워드 기반

Regularized Attentive Capsule Network for Overlapped Relation Extraction

2020-12-18 · COLING 2020 8 · Tianyi Liu, Xiangyu Lin, Weijia Jia, Mingliang Zhou 외

Distantly supervised relation extraction has been widely applied in knowledge base construction due to its less requirement of human efforts. However, the automatically established training datasets in distant supervisio…

DiversityKnowledge Base ConstructionRelationRelation Extraction+1

Orthogonality Constrained Multi-Head Attention For Keyword Spotting

2019-10-10 · Mingu Lee, Jinkyu Lee, Hye Jin Jang, Byeonggeun Kim 외

Multi-head attention mechanism is capable of learning various representations from sequential data while paying attention to different subsequences, e.g., word-pieces or syllables in a spoken word. From the subsequences,…

Keyword Spotting

Aligning Cellular Sheaves with Classifier Attention for Interpretable Weakly-Supervised Pathology Localization

2026-05-24 · Devansh Lalwani, Swapnil Bhat, Maulik Shah arxiv

Weakly-supervised classification of whole-slide images with attention-based multiple instance learning (ABMIL) on top of foundation features now reaches near-saturation on Camelyon16 slide-level performance, but the corr…

Multiple Instance Learning

Scheduled DropHead: A Regularization Method for Transformer Models

2020-04-28 · Findings of the Association for Computational Linguistics 2020 · Wangchunshu Zhou, Tao Ge, Ke Xu, Furu Wei 외

In this paper, we introduce DropHead, a structured dropout method specifically designed for regularizing the multi-head attention mechanism, which is a key component of transformer, a state-of-the-art model for various N…

Machine Translationtext-classificationText ClassificationTranslation

Multi-Head Attention Residuals

2026-07-22 · Cheng Luo, Zefan Cai, Junjie Hu hf

Transformers propagate information across depth through a single additive residual stream: every sublayer reads only the most recent state. Attention residuals relax this by letting each sublayer attend, through a learne…