惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

D
DataBreaches.Net
L
LangChain Blog
博
博客园_首页
J
Java Code Geeks
博
博客园 - 【当耐特】
Microsoft Azure Blog
Microsoft Azure Blog
小众软件
小众软件
WordPress大学
WordPress大学
V
Visual Studio Blog
T
The Blog of Author Tim Ferriss
U
Unit 42
酷 壳 – CoolShell
酷 壳 – CoolShell
Recent Announcements
Recent Announcements
C
Check Point Blog
IT之家
IT之家
Engineering at Meta
Engineering at Meta
N
Netflix TechBlog - Medium
A
About on SuperTechFans
aimingoo的专栏
aimingoo的专栏
D
Docker
有赞技术团队
有赞技术团队
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
阮一峰的网络日志
阮一峰的网络日志
I
InfoQ

cs.LG updates on arXiv.org

Memory-Guided Trust-Region Bayesian Optimization (MG-TuRBO) for High Dimensions EngageTriBoost: Predictive Modeling of User Engagement in Digital Mental Health Intervention Using Explainable Machine Learning Reservoir observer enhanced with residual calibration and attention mechanism Efficient RL Training for LLMs with Experience Replay Wireless Communication Enhanced Value Decomposition for Multi-Agent Reinforcement Learning Adversarial Sensor Errors for Safe and Robust Wind Turbine Fleet Control IKKA: Inversion Classification via Critical Anomalies for Robust Visual Servoing Adaptive Simulation Experiment for LLM Policy Optimization EvoLen: Evolution-Guided Tokenization for DNA Language Model Smartwatch-Based Sitting Time Estimation in Real-World Office Settings Structural Evaluation Metrics for SVG Generation via Leave-One-Out Analysis Loom: A Scalable Analytical Neural Computer Architecture Spectral Geometry of LoRA Adapters Encodes Training Objective and Predicts Harmful Compliance Finite-Sample Analysis of Nonlinear Independent Component Analysis:Sample Complexity and Identifiability Bounds How does Chain of Thought decompose complex tasks? Uncertainty-Aware Transformers: Conformal Prediction for Language Models Adaptive Candidate Point Thompson Sampling for High-Dimensional Bayesian Optimization Using Synthetic Data for Machine Learning-based Childhood Vaccination Prediction in Narok, Kenya Delve into the Applicability of Advanced Optimizers for Multi-Task Learning Bridging SFT and RL: Dynamic Policy Optimization for Robust Reasoning Multi-Agent Decision-Focused Learning via Value-Aware Sequential Communication Predictive Entropy Links Calibration and Paraphrase Sensitivity in Medical Vision-Language Models Efficient Hierarchical Implicit Flow Q-learning for Offline Goal-conditioned Reinforcement Learning Modality-Aware Zero-Shot Pruning and Sparse Attention for Efficient Multimodal Edge Inference The nextAI Solution to the NeurIPS 2023 LLM Efficiency Challenge Feature-Label Modal Alignment for Robust Partial Multi-Label Learning Integrated electro-optic attention nonlinearities for transformers Toward World Models for Epidemiology Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection Batch Distillation Data for Developing Machine Learning Anomaly Detection Methods
RetroInfer: A Vector Storage Engine for Scalable Long-Con...
Yaoqi Chen, · 2026-04-28 · via cs.LG updates on arXiv.org

View PDF

Abstract:Recent large language models (LLMs) are rapidly extending their context windows, yet inference throughput lags due to increasing GPU memory and bandwidth demands. This is because the key-value (KV) cache, an intermediate structure storing token representations, grows linearly with context length and requires an iterative linear scan for attention computation. A promising direction to accelerate long-context inference is to exploit attention's inherent sparsity by offloading the KV cache to CPU memory and retrieving only a small subset of tokens important to the current generation step. However, prior sparse attention approaches struggle to balance accuracy and retrieval cost due to varying sparsity patterns and inefficient GPU-CPU memory management.
We present RetroInfer, a vector storage engine that realizes a sparsity-based KV cache for long-context inference. RetroInfer introduces an Attention-aWare VEctor index (wave index), which fundamentally improves the tradeoff between attention accuracy and retrieval cost through tripartite attention approximation, accuracy-bound attention estimation, and segmented clustering. We also design the wave buffer, a GPU-CPU buffer manager that assigns computation and manages data across heterogeneous hardware. We evaluate RetroInfer across a range of models and workloads, demonstrating up to 4.4X decoding throughput over full attention at 120K context and up to 12.2X over sparse attention baselines at 1 million tokens -- all while preserving full-attention-level accuracy.
Comments: 16 pages; Accepted by VLDB 2026
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2505.02922 [cs.LG]
  (or arXiv:2505.02922v3 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2505.02922

arXiv-issued DOI via DataCite
Journal reference: PVLDB, 19(5): 1016-1031, 2026
Related DOI: https://doi.org/10.14778/3796195.3796212

DOI(s) linking to related resources

Submission history

From: Baotong Lu [view email]
[v1] Mon, 5 May 2025 18:01:17 UTC (676 KB)
[v2] Mon, 30 Jun 2025 05:21:58 UTC (681 KB)
[v3] Mon, 27 Apr 2026 10:13:35 UTC (600 KB)

此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。