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

U
Unit 42
L
LangChain Blog
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Y
Y Combinator Blog
钛媒体：引领未来商业与生活新知
钛媒体：引领未来商业与生活新知
J
Java Code Geeks
有赞技术团队
有赞技术团队
B
Blog
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
V
V2EX
Apple Machine Learning Research
Apple Machine Learning Research
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
WordPress大学
WordPress大学
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
阮一峰的网络日志
阮一峰的网络日志
The Cloudflare Blog
Martin Fowler
Martin Fowler
H
Hackread – Cybersecurity News, Data Breaches, AI and More
M
MIT News - Artificial intelligence
Recent Announcements
Recent Announcements
D
DataBreaches.Net
The GitHub Blog
The GitHub Blog
博
博客园 - Franky
小众软件
小众软件

cs.DC updates on arXiv.org

DUAL-BLADE: Dual-Path NVMe-Direct KV-Cache Offloading for Edge LLM Inference Progressive Semantic Communication for Efficient Edge-Cloud Vision-Language Models Efficient, VRAM-Constrained xLM Inference on Clients Folding Tensor and Sequence Parallelism for Memory-Efficient Transformer Training & Inference DORA: A Scalable Asynchronous Reinforcement Learning System for Language Model Training AMMA: A Multi-Chiplet Memory-Centric Architecture for Low-Latency 1M Context Attention Serving RaMP: Runtime-Aware Megakernel Polymorphism for Mixture-of-Experts Spark Policy Toolkit: Semantic Contracts and Scalable Execution for Policy Learning in Spark Internet of Everything in the 6G Era: Paradigms, Enablers, Potentials and Future Directions PolyKV: A Shared Asymmetrically-Compressed KV Cache Pool for Multi-Agent LLM Inference A Survey on Split Learning for LLM Fine-Tuning: Models, Systems, and Privacy Optimizations ITAS: A Multi-Agent Architecture for LLM-Based Intelligent Tutoring Latency and Cost of Multi-Agent Intelligent Tutoring at Scale TACO: Efficient Communication Compression of Intermediate Tensors for Scalable Tensor-Parallel LLM Training FreeScale: Distributed Training for Sequence Recommendation Models with Minimal Scaling Cost CommFuse: Hiding Tail Latency via Communication Decomposition and Fusion for Distributed LLM Training A Taxonomy and Resolution Strategy for Client-Level Disagreements in Federated Learning Usable Agent Discovery for Decentralized AI Systems Cloud to Edge: Benchmarking LLM Inference On Hardware-Accelerated Single-Board Computers Data-Free Contribution Estimation in Federated Learning using Gradient von Neumann Entropy Shard the Gradient, Scale the Model: Serverless Federated Aggregation via Gradient Partitioning Promoting Simple Agents: Ensemble Methods for Event-Log Prediction GraphLeap: Decoupling Graph Construction and Convolution for Vision GNN Acceleration on FPGA AGNT2: Autonomous Agent Economies on Interaction-Optimized Layer 2 Infrastructure FedSIR: Spectral Client Identification and Relabeling for Federated Learning with Noisy Labels Stream-CQSA: Avoiding Out-of-Memory in Attention Computation via Flexible Workload Scheduling A Delta-Aware Orchestration Framework for Scalable Multi-Agent Edge Computing Federated Learning over Blockchain-Enabled Cloud Infrastructure Optimal Routing for Federated Learning over Dynamic Satellite Networks: Tractable or Not? Sherpa.ai Privacy-Preserving Multi-Party Entity Alignment without Intersection Disclosure for Noisy Identifiers
Contextual Chain: A Context-Based Design Principle and Co...
[Submitted on 8 Apr 2026 (v1), last revised 25 Aug 2026 (this ve · 2026-04-08 · via cs.DC updates on arXiv.org

View PDF HTML (experimental)

Abstract:We introduce Contextual Chain as a design principle in which evolving operational context guides subsequent system decisions. This article evaluates a first systems realization for recovery after intermittent connectivity. The central controlled result is that context-informed timing improves recovery even when assigned synchronization quantity is held fixed. In an exact equal-assigned-budget control, Adaptive and FixedMatched receive the same total assigned pair budget for every case and seed, yet the context-triggered Adaptive schedule raises final agreement from 0.802 to 0.864 in Case A and from 0.800 to 0.860 in Case B and shortens mean recovery by 10.8 and 14.0 s relative to approximately uniform timing. Thus total assigned synchronization quantity alone cannot explain Adaptive performance: within the oracle-assisted simulator, contextual inconsistency is an effective control signal for temporal resource allocation. A matched-seed 3 x 3 factorial further shows that synchronization provisioning is the dominant recovery factor, whereas the present fuller head-selection rule has no consistent advantage over height-only selection; BranchScoreOnly is operationally identical to Full under the main height-epoch configuration. The demonstrated effect in this first Contextual Chain realization is therefore context-informed scheduling rather than superiority of the tested fork-choice formula. The broader hypothesis is a resource asymmetry in which synchronized participants follow one current context while a context-uncertain observer may need to retain multiple plausible histories. The present study establishes a systems-level role for context-informed resource allocation, while cryptographic authentication and post-quantum formalization, fully distributed triggering, and deployment-level traffic and energy validation define the next research directions for Contextual Chain.

Submission history

From: Song-Ju Kim Dr. [view email]
[v1] Wed, 8 Apr 2026 00:02:37 UTC (620 KB)
[v2] Tue, 25 Aug 2026 18:09:59 UTC (878 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。