BestHub
Sign in
Tagged articles

Incremental Update

14 articles · Page 1 of 1
ThinkingAgent
ThinkingAgent
Jul 6, 2026 · Artificial Intelligence

Turning Data into Model-Ready Knowledge with RAG Pipelines and Vector DBs

An enterprise RAG pipeline must transform scattered documents into timely, secure, and explainable knowledge for LLMs, covering parsing, cleaning, chunking (recursive, semantic, contextual), embedding with BGE‑M3, hybrid vector‑BM25‑graph retrieval, RRF fusion, cross‑encoder rerank, ACL pre‑filtering, and minute‑level incremental updates.

ACLEmbeddingHybrid Retrieval
0 likes · 28 min read
Turning Data into Model-Ready Knowledge with RAG Pipelines and Vector DBs
PaperAgent
PaperAgent
May 18, 2026 · Artificial Intelligence

How MemWeaver Combines Behavioral and Cognitive Memory to Rebuild LLM Personalization

MemWeaver introduces a hierarchical memory that fuses behavior‑level and cognition‑level user signals, enabling large language models to generate more personalized content across multiple tasks, with extensive experiments, ablations, and an efficient incremental update mechanism demonstrating superior performance over strong baselines.

Incremental UpdateLLM personalizationLaMP benchmark
0 likes · 12 min read
How MemWeaver Combines Behavioral and Cognitive Memory to Rebuild LLM Personalization
AI Algorithm Path
AI Algorithm Path
May 22, 2025 · Artificial Intelligence

Monte Carlo Policy Improvement in RL: Epsilon‑Greedy, On‑Policy vs Off‑Policy, and Incremental Updates

This tutorial explains how Monte Carlo methods are enhanced in reinforcement learning through epsilon‑greedy and epsilon‑soft policies, Monte Carlo control, a Blackjack Q‑function example, the distinction between on‑policy and off‑policy learning, importance sampling, and efficient incremental update techniques.

Epsilon-GreedyImportance SamplingIncremental Update
0 likes · 14 min read
Monte Carlo Policy Improvement in RL: Epsilon‑Greedy, On‑Policy vs Off‑Policy, and Incremental Updates
AI Large Model Application Practice
AI Large Model Application Practice
Oct 30, 2024 · Artificial Intelligence

How to Efficiently Incrementally Update Knowledge in RAG Applications

Incremental knowledge updates in Retrieval‑Augmented Generation (RAG) systems can be achieved by using document‑level or chunk‑level strategies, leveraging hash fingerprints, record managers, and framework‑specific APIs such as LangChain’s index() with cleanup modes or LlamaIndex’s ingestion pipeline, reducing redundant computation and cost.

Incremental UpdateLangChainRAG
0 likes · 12 min read
How to Efficiently Incrementally Update Knowledge in RAG Applications
Su San Talks Tech
Su San Talks Tech
Apr 22, 2024 · Fundamentals

Why Java’s Object Graph Traversal Pauses: Tri‑Color Marking & STW Solutions

This article explains the two‑phase reachability analysis in Java garbage collection, detailing why the object‑graph traversal stage incurs stop‑the‑world pauses, how the tri‑color marking model reveals issues like floating garbage and object disappearance, and describes incremental update and snapshot‑at‑the‑beginning techniques that avoid STW.

Garbage CollectionIncremental UpdateSATB
0 likes · 14 min read
Why Java’s Object Graph Traversal Pauses: Tri‑Color Marking & STW Solutions
NetEase Cloud Music Tech Team
NetEase Cloud Music Tech Team
Jan 24, 2024 · Mobile Development

Optimizing Hermes Bytecode Bundle Size and SourceMap Handling in React Native

To counter the 40‑100 % size increase of Hermes bytecode bundles after upgrading to React Native 0.70, the team switched to xz compression (cutting zip size 20‑26 %), enabled hermesc ‑O optimization (shrinking bundles 10‑22 % while requiring a two‑step source‑map merge), and used the ‑base‑bytecode option with bsdiff to reduce incremental OTA patches by up to 85 %, collectively improving download and update efficiency.

BytecodeHermesIncremental Update
0 likes · 13 min read
Optimizing Hermes Bytecode Bundle Size and SourceMap Handling in React Native
DataFunTalk
DataFunTalk
Nov 12, 2023 · Big Data

MaxCompute Incremental Update Architecture, Intelligent Materialized Views, and Adaptive Execution Optimizations

This article presents a comprehensive overview of MaxCompute's near‑real‑time incremental update and processing architecture, the design of Transactional Table 2.0, intelligent materialized view evolution and recommendation, as well as multi‑level adaptive execution optimizations for the SQL engine, illustrating how these innovations improve efficiency, cost, and scalability for large‑scale data workloads.

Adaptive ExecutionIncremental UpdateMaxCompute
0 likes · 20 min read
MaxCompute Incremental Update Architecture, Intelligent Materialized Views, and Adaptive Execution Optimizations
Alipay Experience Technology
Alipay Experience Technology
Sep 19, 2022 · Frontend Development

How Yuque Built a Secure, High‑Performance Desktop App with Electron

This article details Yuque's desktop client architecture, covering its multi‑layer design, security measures, incremental update strategy, reusable components, delivery efficiency, performance optimizations, and stability practices that enable a fast, offline‑capable, and user‑friendly productivity tool.

Desktop applicationElectronIncremental Update
0 likes · 15 min read
How Yuque Built a Secure, High‑Performance Desktop App with Electron
Shopee Tech Team
Shopee Tech Team
Mar 10, 2022 · Mobile Development

FolderBsdp: An Efficient Incremental Update Solution for React Native Hybrid Apps

FolderBsdp provides an efficient incremental update method for React Native hybrid apps by diffing raw resource folders per file with BSDiff, packaging tiny patches into a single ZIP, dramatically cutting bandwidth, memory usage, and storage while eliminating extra ZIP dependencies, making it ideal for low‑end devices.

Diff AlgorithmFolderBsdpIncremental Update
0 likes · 22 min read
FolderBsdp: An Efficient Incremental Update Solution for React Native Hybrid Apps
Meituan Technology Team
Meituan Technology Team
Jul 5, 2018 · Backend Development

Incremental Update Practices and Go‑Based Performance Optimization at Meituan Finance

Meituan Finance reduced user churn by replacing Node‑based diff generation with a Go‑powered incremental‑update service that uses lightweight goroutines, a three‑layer CDN/API/compute architecture, and pre‑warmed patches, achieving 99.97 % success, 64.9 % incremental updates and roughly 164 KB daily bandwidth savings per user.

FrontendGoIncremental Update
0 likes · 19 min read
Incremental Update Practices and Go‑Based Performance Optimization at Meituan Finance
Ctrip Technology
Ctrip Technology
Feb 28, 2018 · Mobile Development

Incremental Offline Package Update Solution for Ctrip Travel App

The article details Ctrip's engineering solution for incremental offline package updates in its travel app, covering package creation, management, differential patch generation, download strategies, installation, monitoring, and practical challenges to achieve high success rates on iOS and Android.

Diff AlgorithmIncremental Updatedeployment
0 likes · 13 min read
Incremental Offline Package Update Solution for Ctrip Travel App
Tencent Music Tech Team
Tencent Music Tech Team
Jun 6, 2016 · Mobile Development

Incremental Update Solution for Android APKs Using Binary Diff and Channel Management

The article presents an Android incremental‑update solution that uses bsdiff‑generated binary patches to replace full APK downloads, embeds channel identifiers in the ZIP comment field to avoid per‑channel APK duplication, and details server‑side patch creation, client‑side channel stripping, patch merging, and verification to reduce bandwidth and simplify multi‑channel distribution.

APK PatchAndroidIncremental Update
0 likes · 18 min read
Incremental Update Solution for Android APKs Using Binary Diff and Channel Management