paper-with-me

홈 › Papers

Representation Sparsification with Hybrid Thresholding for Fast SPLADE-based Document Retrieval

2023-06-20 · Yifan Qiao, Yingrui Yang, Shanxiu He, Tao Yang

Learned sparse document representations using a transformer-based neural model has been found to be attractive in both relevance effectiveness and time efficiency. This paper describes a representation sparsification scheme based on hard and soft thresholding with an inverted index approximation for faster SPLADE-based document retrieval. It provides analytical and experimental results on the impact of this learnable hybrid thresholding scheme.

📄 PDF Abstract BibTeX arXiv:2306.11293

Code (1)

qiaoyf96/ht 공식 구현

Tasks

Retrieval

Similar Papers 제목 키워드 기반

CHESS: Optimizing LLM Inference via Channel-Wise Thresholding and Selective Sparsification

2024-09-02 · Junhui He, Shangyu Wu, Weidong Wen, Chun Jason Xue 외

Deploying large language models (LLMs) on edge devices presents significant challenges due to the substantial computational overhead and memory requirements. Activation sparsification can mitigate these resource challeng…

An Alternative to FLOPS Regularization to Effectively Productionize SPLADE-Doc

2025-05-21 · Aldo Porco, Dhruv Mehra, Igor Malioutov, Karthik Radhakrishnan 외

Learned Sparse Retrieval (LSR) models encode text as weighted term vectors, which need to be sparse to leverage inverted index structures during retrieval. SPLADE, the most popular LSR model, uses FLOPS regularization to…

Retrieval

Exact Hybrid Covariance Thresholding for Joint Graphical Lasso

2015-03-07 · Qingming Tang, Chao Yang, Jian Peng, Jinbo Xu

This paper considers the problem of estimating multiple related Gaussian graphical models from a $p$-dimensional dataset consisting of different classes. Our work is based upon the formulation of this problem as group gr…

Exploring the Representation Power of SPLADE Models

2023-06-29 · Joel Mackenzie, Shengyao Zhuang, Guido Zuccon

The SPLADE (SParse Lexical AnD Expansion) model is a highly effective approach to learned sparse retrieval, where documents are represented by term impact scores derived from large language models. During training, SPLAD…

Retrieval

The Role of Vocabularies in Learning Sparse Representations for Ranking

2025-09-20 · Hiun Kim, Tae Kwan Lee, Taeryun Won arxiv

Learned Sparse Retrieval (LSR) such as SPLADE has growing interest for effective semantic 1st stage matching while enjoying the efficiency of inverted indices. A recent work on learning SPLADE models with expanded vocabu…