Research Papers
Research Papers news — Thursday 15 October 2026
26 stories from 3 sources, every one in English. Each links to its publisher.
-
Atomic functional site scaffolding in biopolymers with RFdiffusion3
Science Magazine Tech
Science, Ahead of Print.
-
Magic state cultivation on a superconducting quantum processor
Science Magazine Tech
Science, Ahead of Print.
-
Computational compressive imaging reconstruction with a simulated-annealing-based Ising solver
arXiv cs.DC (Distributed)
arXiv:2610.19778v1 Announce Type: cross Abstract: Computational compressive imaging, such as single-pixel imaging and lensless imaging, is a promising imaging technique, enabling the reconstruction of objects from under-sampled…
-
Delta Tokenizer: Reference-Identical Tokenization of Session Deltas and Segment-Parallelism for Long Prompts on Multicore CPUs
arXiv cs.DC (Distributed)
arXiv:2610.19801v1 Announce Type: cross Abstract: An agent that drives an LLM sends the model the same conversation again and again, and each turn adds only a little text at the end. Inference engines reuse the KV cache of the shared…
-
Real-Time GPU-Accelerated Monte Carlo Evaluation of Safety-Critical AEB Systems Under Uncertainty
arXiv cs.DC (Distributed)
arXiv:2604.27193v2 Announce Type: replace-cross Abstract: Automatic Emergency Braking (AEB) systems represent a safety-critical national interest, with the National Highway Traffic Safety Administration (NHTSA) Federal Motor Vehicle Safety…
-
TrimSO(2): Fast and Memory-Efficient Training of Equivariant Interatomic Potentials
arXiv cs.DC (Distributed)
arXiv:2610.20029v1 Announce Type: new Abstract: Equivariant interatomic potentials enable accurate and efficient simulations of molecules and materials. Among them, models based on SO(2) convolutions reduce the cost of tensor products and…
-
JointServe: Serving Multi-Model Agents on Memory-Constrained GPUs
arXiv cs.DC (Distributed)
arXiv:2610.19822v1 Announce Type: new Abstract: LLM agents are evolving from a single model calling tools to planners that orchestrate multiple expert models whose capabilities live in distinct weights. This shift breaks a hidden…
-
Beyond the Trial-and-Error Loop: Hybrid Projection and Automated Tuning for Distributed Training
arXiv cs.DC (Distributed)
arXiv:2610.19423v1 Announce Type: cross Abstract: Configuring distributed LLM training across hundreds of GPUs requires jointly optimizing tensor, pipeline, expert, context, and data parallelism alongside micro-batch size, recomputation…
-
How the Dirty Set Moves: Characterizing Schedule Reuse in Incremental Sparse Matrix Multiplication
arXiv cs.DC (Distributed)
arXiv:2610.19489v1 Announce Type: cross Abstract: Sparse matrices, in which most entries are zero, are central to scientific simulation, graph analytics and machine learning, and their multiplication with dense inputs is a core parallel…
-
Konflux: Optimized Function Fusion for Serverless Applications
arXiv cs.DC (Distributed)
arXiv:2601.11156v2 Announce Type: replace Abstract: Function-as-a-Service (FaaS) has become a central paradigm in serverless cloud computing, yet optimizing FaaS deployments remains challenging. Using function fusion, multiple functions…
-
CacheSmith: Accurate RAG Serving under Latency SLOs with Adaptive KV Reuse
arXiv cs.DC (Distributed)
arXiv:2610.18925v1 Announce Type: cross Abstract: This paper presents CacheSmith, a KV cache reuse system for RAG prefill that maximizes quality-gated SLO attainment. While context caching lowers prefill latency by reusing precomputed KV…
-
Prioritizing Microservice Requests via Retries
arXiv cs.DC (Distributed)
arXiv:2610.20567v1 Announce Type: cross Abstract: Cloud services rapidly shift towards independently-managed microservices architectures. This independence may lead to temporary microservices miscoordination, resulting in overloads and…
-
The Cost of Context: Profiling the Energy Footprint of Input Tokens in Large Language Models
arXiv cs.DC (Distributed)
arXiv:2610.19819v1 Announce Type: new Abstract: The rapid adoption of Retrieval-Augmented Generation (RAG) and long-context LLMs has shifted the computational burden toward the prefill phase---the initial stage where the model processes…
-
Tracing Knowledge Lineage for Federated Unlearning under Mutual Distillation
arXiv cs.DC (Distributed)
arXiv:2610.19072v1 Announce Type: cross Abstract: An edge server and nearby devices can train language models of different sizes together by mutual distillation. Federated unlearning (FU) and unlearning for large language models (LLMs)…
-
Dynamic 3-Coloring of Bipartite Graphs with Optimal Locality
arXiv cs.DC (Distributed)
arXiv:2610.19710v1 Announce Type: cross Abstract: We give a dynamic graph algorithm which maintains a 3-coloring in bipartite graphs under edge insertions such that the coloring only changes within a radius of $O(\log n)$ around an…
-
Arcturus: A Cloud Overlay Network for Global Accelerator with Enhanced Performance and Stability
arXiv cs.DC (Distributed)
arXiv:2507.10928v2 Announce Type: replace-cross Abstract: Global Accelerator (GA) services play a vital role in ensuring low-latency, high-reliability communication for real-time interactive applications. However, existing GA offerings are…
-
Parallel Sparse Outputs for Tensor Compilers
arXiv cs.DC (Distributed)
arXiv:2610.19308v1 Announce Type: new Abstract: Sparse tensor compilers are commonly used high-performance tools that optimize tensor algebra programs over notoriously complex sparse data structures. Despite the ever-increasing size and…
-
Scaling LLM Inference via Parallelism: Throughput-Latency Trade-off
arXiv cs.DC (Distributed)
arXiv:2610.19405v1 Announce Type: new Abstract: Large Language Model (LLM) serving systems must balance latency and throughput while efficiently utilizing distributed GPU resources. This challenge arises because inference consists of two…
-
Federated Unlearning of Quantum Data via Classical Shadows
arXiv cs.DC (Distributed)
arXiv:2610.19077v1 Announce Type: cross Abstract: Quantum federated learning trains a quantum neural network (QNN) across nodes that hold quantum data, such as the states of quantum sensors. To forget data, unlearning methods for quantum…
-
sGLP: Simulating People with Stochastic Agents
arXiv cs.DC (Distributed)
arXiv:2610.14738v2 Announce Type: replace-cross Abstract: Grassroots platforms are designed to be owned, operated and governed by their participants, and thus are constituted from person-to-person smartphone apps that are run by the…
-
Native Dual Pools or Shared Paging? Memory Organization for Concurrent Hybrid LLM Serving
arXiv cs.DC (Distributed)
arXiv:2610.15755v2 Announce Type: replace Abstract: Hybrid attention-SSM serving combines growing attention KV with fixed-size recurrent and convolution state whose allocation units, reuse boundaries, and lifetimes differ. We implement…
-
Substep Linearizability: Relaxing Atomicity While Preserving Sequential Specification
arXiv cs.DC (Distributed)
arXiv:2610.16099v2 Announce Type: replace Abstract: Linearizability is a standard correctness condition for concurrent objects. It takes a correct by construction sequential specification of a concurrent data structure, and produces a…
-
Understanding NVIDIA Grace-Blackwell Unified Memory Through Placement-Controlled Microbenchmarks
arXiv cs.DC (Distributed)
arXiv:2610.20354v1 Announce Type: new Abstract: NVIDIA Grace--Blackwell systems make CPU--GPU memory increasingly unified at the programming level, but unified addressability does not make the memory system uniform. A CUDA managed…
-
Rethinking Sparse Formats for RISC-V: A Hierarchical Approach to High-Performance SpMV
arXiv cs.DC (Distributed)
arXiv:2609.11352v2 Announce Type: replace Abstract: The sparse matrix-vector multiplication (SpMV) algorithm is a fundamental computational kernel of linear algebra and serves as a building block for numerous applications, primarily…
-
A Self-Defining Operator for Distributed Applications
arXiv cs.DC (Distributed)
arXiv:2610.14547v2 Announce Type: replace Abstract: To operate distributed applications with high availability, operators must rapidly diagnose and mitigate any incident. LLM-based agents can help, but unlike experienced human operators…
-
Is AI eroding PhD students’ critical thinking and analysis skills?
Nature Technology