Research Papers
Research Papers news — Wednesday 21 October 2026
29 stories from 1 sources, every one in English. Each links to its publisher.
-
Specter: Fast MoE Inference via Quantization-Enabled Multi-Step Speculative Offloading
arXiv cs.DC (Distributed)
arXiv:2511.14102v2 Announce Type: replace-cross Abstract: Mixture-of-experts (MoE) models often exceed a single accelerator's memory capacity. Offloading experts to host memory creates a severe PCIe I/O bottleneck because data-dependent…
-
Discount Yourself: Self-inclusive Freshness-weighted Gossip for Decentralised Learning under Client Outages
arXiv cs.DC (Distributed)
arXiv:2610.28545v1 Announce Type: cross Abstract: Decentralised federated learning (DFL) trains personalised models on devices that average with their neighbours rather than with a central server. Real devices, however, are often offline…
-
Fault Tolerance for Stream Processing Engines
arXiv cs.DC (Distributed)
arXiv:1605.00928v4 Announce Type: replace Abstract: Distributed Stream Processing Engines (DSPEs) target applications related to continuous computation, online machine learning and real-time query processing. DSPEs operate on high volume…
-
LLM-Attack Suite: A Unified Red-Team Framework for Privacy Evaluation of Split and Distributed LLM Systems
arXiv cs.DC (Distributed)
arXiv:2610.22722v2 Announce Type: replace-cross Abstract: Split language models keep input text on a client but send intermediate representations to a server. Attacks can reconstruct text from these messages. Comparing privacy evaluations…
-
From Last-Layer to Cross-Layer: One-Shot Joint Classifier Reconstruction for Federated Learning
arXiv cs.DC (Distributed)
arXiv:2610.26953v1 Announce Type: new Abstract: Federated learning (FL) enables collaborative model training while preserving data privacy. However, non-independent and identically distributed (non-IID) data inevitably introduces…
-
Emulating DGEMM and ZGEMM with INT8 Matrix Engines on Intel Arc GPUs: Ozaki Scheme II on the Xe2 Architecture
arXiv cs.DC (Distributed)
arXiv:2610.26945v1 Announce Type: cross Abstract: Double-precision matrix multiplication, real (DGEMM) and complex (ZGEMM), dominates the run time of dense linear algebra and electronic-structure codes. On affordable GPUs, however, FP64…
-
BitMTP: When Multi-Token Prediction Meets Low-Bit Large Language Models
arXiv cs.DC (Distributed)
arXiv:2610.28500v1 Announce Type: cross Abstract: For efficient deployment of large language models (LLMs), methods of ultra-low-bit backbones reduce weight storage, memory traffic, and linear-layer execution cost, while multi-token…
-
A mechanism for collective foraging: percolating recruitment cascades coupled to a maximum-entropy
arXiv cs.DC (Distributed)
arXiv:2610.24865v2 Announce Type: replace-cross Abstract: Foraging recruitment is a contagion: an agent that finds a resource triggers a linear-threshold cascade, the model built for rumour and adoption spread on social networks, now…
-
Weight Placement Is Not Enough: Synchronization in Multi-Node CPU Decoding of Ternary LLMs
arXiv cs.DC (Distributed)
arXiv:2610.28620v1 Announce Type: new Abstract: Ternary LLMs read only a few gigabytes per decoded token, yet on dual-socket servers with four NUMA nodes the fastest ternary runtime we found uses less than half of the memory bandwidth…
-
SweepANN: A GPU-Accelerated Large-K Approximate Vector Search for Drug Discovery
arXiv cs.DC (Distributed)
arXiv:2610.27306v1 Announce Type: cross Abstract: DrugCLIP places small molecules and protein pockets in one space, turning genome-wide drug discovery into large-K retrieval by vector proximity. A campaign fixes over 20K pockets in advance…
-
Magneton: Optimizing Energy Efficiency of ML Systems via Differential Energy Debugging
arXiv cs.DC (Distributed)
arXiv:2512.08365v2 Announce Type: replace Abstract: The training and deployment of machine learning (ML) models have become extremely energy-intensive. While existing optimization efforts focus primarily on hardware energy efficiency, a…
-
DIFFLET: A Fast Inference Engine for Diffusion Transformer Models on AWS Trainium
arXiv cs.DC (Distributed)
arXiv:2610.23047v2 Announce Type: replace Abstract: On AWS Trainium2, every component of a diffusion model runs as a program compiled ahead of time for fixed shapes, while pipelines differ across models, prompt lengths across requests, and…
-
Learning-Theoretic Foundation for General Coded Computing: The Adversarial Setting
arXiv cs.DC (Distributed)
arXiv:2610.27440v1 Announce Type: cross Abstract: Adversarial workers pose a severe challenge in distributed computing systems by returning corrupted outputs that can substantially distort the recovered computations. Coded computing…
-
HyCo: Hybrid Drafting for Cost-Efficient Mobile-Cloud Speculative Decoding
arXiv cs.DC (Distributed)
arXiv:2610.21436v2 Announce Type: replace-cross Abstract: Device-cloud speculative decoding reduces server computation, but growing a deep draft tree with a small language model (SLM) requires repeated passes on a mobile neural processing…
-
Evaluating Runtime Prediction and Conservative Walltime Policies in HPC Systems
arXiv cs.DC (Distributed)
arXiv:2610.27755v1 Announce Type: new Abstract: Efficient use of large-scale supercomputing systems depends on accurate runtime estimation and walltime selection. To avoid premature termination, users often overestimate walltime, leading…
-
PULSE: Personalized Opportunistic Learning for Wearable Devices
arXiv cs.DC (Distributed)
arXiv:2610.28081v1 Announce Type: cross Abstract: Wearable devices increasingly support personalized health and well-being applications, but post-deployment data can differ substantially from those used for training, while purely local…
-
QwenGyre: An Elastic Reinforcement Learning Framework for Training xLong-Horizon Agents
arXiv cs.DC (Distributed)
arXiv:2609.33848v2 Announce Type: replace-cross Abstract: Large language model (LLM) agents increasingly undertake extreme-long (xlong) horizon tasks, where a single execution can span hours, hundreds of model--environment interactions…
-
VS-Flash: Harnessing Vector Search with Computational Memory-Semantic Flash
arXiv cs.DC (Distributed)
arXiv:2610.26960v1 Announce Type: cross Abstract: As generative AI continues to evolve, retrieval-augmented generation (RAG) systems are driving rapid growth in vector database capacity requirements. As datasets outgrow host DRAM, vector…
-
Unsupervised Head Network Personalization for Split Computing via Federated Learning
arXiv cs.DC (Distributed)
arXiv:2610.27952v1 Announce Type: cross Abstract: Split Computing (SC) is a distributed inference method designed for balancing computational load and reducing latency by partitioning a neural network into head and tail networks and…
-
Online Continual Learning on Intel Loihi 2 via a Co-designed Spiking Neural Network
arXiv cs.DC (Distributed)
arXiv:2511.01553v3 Announce Type: replace-cross Abstract: Edge AI must adapt to new classes and shifting data within a tight power budget, but online continual learning (learning from each sample before the next arrives) is where…
-
The Token-Energy Law of LLM Decode: Active Bytes and Realized Time Account for Joules per Token
arXiv cs.DC (Distributed)
arXiv:2610.28576v1 Announce Type: cross Abstract: Energy-efficient deployment of large language models requires understanding how model architecture, hardware, and serving settings shape energy per generated token. We introduce the…
-
Budgeted Admission from Timestamped Capacity Records in Shared Clusters
arXiv cs.DC (Distributed)
arXiv:2610.28392v1 Announce Type: new Abstract: A cluster admission gate can receive a capacity record that no longer matches current schedulability. We study one exogenous binary schedulability unit under a bound on the joint probability…
-
3D-MoE: Complementarity-Guided Hardware-Software Co-Design for MoE Decode on 3D-Stacked NMP
arXiv cs.DC (Distributed)
arXiv:2610.26991v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) have become a mainstream way to scale Large Language Models (LLMs), because each token activates only a small subset of experts, reducing inference cost. However…
-
NVIDIA Spectrum-X: Hardware-Accelerated Multiplane Networking for Large-Scale AI Training
arXiv cs.DC (Distributed)
arXiv:2605.21187v2 Announce Type: replace-cross Abstract: As distributed model training scales to span hundreds of thousands of GPUs, scale-out networks face unprecedented performance and efficiency demands. NVIDIA Spectrum-X Ethernet has…
-
FEAST: Federated Shared-Space Training for Resource-Heterogeneous Clients
arXiv cs.DC (Distributed)
arXiv:2608.09250v2 Announce Type: replace-cross Abstract: Federated learning (FL) must serve devices with varying computational capabilities. A fixed model cannot suit all devices, while training one model per deployment limit is costly…
-
Benchmarking CPU, GPU, and ML Power-Flow Solvers from 14 to 70,000 Buses
arXiv cs.DC (Distributed)
arXiv:2610.27930v1 Announce Type: new Abstract: Learned surrogates for power-flow (PF) computation are usually motivated by speed, compared against a single CPU-based baseline. We show this comparison is fragile: across 7 open-source…
-
KPage: Cross-Block Page Compaction for Hardware-Efficient Sparse Attention
arXiv cs.DC (Distributed)
arXiv:2610.28096v1 Announce Type: new Abstract: Sparse attention reduces the computational cost of long-sequence video Transformers, but three-dimensional token layouts often leave selected blocks partially occupied, wasting physical tile…
-
Execution-Time Validation of Context-Carrying Action Intents in Industrial Cyber-Physical Systems
arXiv cs.DC (Distributed)
arXiv:2610.27971v1 Announce Type: cross Abstract: In industrial cells, commands from learned policies, operators and schedulers reach the controller after inference, confirmation and transport, and the plant may have changed meanwhile, for…
-
Constrained Federated Reinforcement Learning for Secure Offloading in Vehicular Edge Computing
arXiv cs.DC (Distributed)
arXiv:2610.27651v1 Announce Type: new Abstract: Vehicular Edge Computing (VEC) reduces latency for computation-intensive vehicular applications by offloading tasks to roadside edge servers. Real vehicular workloads, however, consist of…