paper-with-me

Papers

Sparse and Structured Visual Attention

2020-02-13 · Pedro Henrique Martins, Vlad Niculae, Zita Marinho, André Martins

Visual attention mechanisms are widely used in multimodal tasks, as visual question answering (VQA). One drawback of softmax-based attention mechanisms is that they assign some probability mass to all image regions, regardless of their adjacency structure and of their relevance to the text. In this paper, to better link the image structure with the text, we replace the traditional softmax attention mechanism with two alternative sparsity-promoting transformations: sparsemax, which is able to select only the relevant regions (assigning zero weight to the rest), and a newly proposed Total-Variation Sparse Attention (TVmax), which further encourages the joint selection of adjacent spatial locations. Experiments in VQA show gains in accuracy as well as higher similarity to human attention, which suggests better interpretability.

📄 PDF Abstract BibTeX arXiv:2002.05556

Code (1)

deep-spin/TVmax 공식 구현 pytorch

Tasks

Image CaptioningQuestion AnsweringVisual Question AnsweringVisual Question Answering (VQA)

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Position-Wise Feed-Forward Layer 설명 없음
Sigmoid Activation 설명 없음
Tanh Activation 설명 없음
Residual Connection 설명 없음
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…

Similar Papers 제목 키워드 기반

Compact Attention: Exploiting Structured Spatio-Temporal Sparsity for Fast Video Generation

2025-08-18 · Qirui Li, Guangcong Zheng, Qi Zhao, Jie Li 외 arxiv

The computational demands of self-attention mechanisms pose a critical challenge for transformer-based video generation, particularly in synthesizing ultra-long sequences. Current approaches, such as factorized attention…

Video Generation

Sparse Forcing: Native Trainable Sparse Attention for Real-time Autoregressive Diffusion Video Generation

2026-04-23 · Boxun Xu, Yuming Du, Zichang Liu, Siyu Yang 외 arxiv

We introduce Sparse Forcing, a training-and-inference paradigm for autoregressive video diffusion models that improves long-horizon generation quality while reducing decoding latency. Sparse Forcing is motivated by an em…

Text-to-Video Generation

Sparse Fuzzy Attention for Structured Sentiment Analysis

2021-09-14 · Letian Peng, Zuchao Li, Hai Zhao

Attention scorers have achieved success in parsing tasks like semantic and syntactic dependency parsing. However, in tasks modeled into parsing, like structured sentiment analysis, "dependency edges" are very sparse whic…

Dependency ParsingSentiment Analysis

Interpretable and Sparse Linear Attention with Decoupled Membership-Subspace Modeling via MCR2 Objective

2026-01-20 · Tianyuan Liu, Libin Hou, Linyuan Wang, Bin Yan arxiv

Maximal Coding Rate Reduction (MCR2)-driven white-box transformer, grounded in structured representation learning, unifies interpretability and efficiency, providing a reliable white-box solution for visual modeling. How…

Computational EfficiencyRepresentation Learning

When Structured Sparse Autoencoders Learn Consistent Concepts Across Modalities

2026-07-09 · Weiduo Liao, Yunqiao Yang, Ying Wei arxiv

Sparse autoencoders (SAEs) have emerged as a promising technique for mechanistic interpretability by learning a set of sparse latent features in large models, each of which encodes a distinct concept. However, in vision-…