Research Papers
Research Papers news — Monday 28 September 2026
38 stories from 2 sources, every one in English. Each links to its publisher.
-
Quantum-HPC Workflows Across Multiple Quantum Computing Platforms: Two Case Studies
arXiv cs.DC (Distributed)
arXiv:2609.31528v1 Announce Type: cross Abstract: We describe the Quantum-HPC hybrid system comprising the supercomputer Fugaku, the QPUs Quantinuum Reimei and ibm_kobe, and the Tierkreis hybrid workflow software. We present two…
-
Encryptability As a Coordinate Choice: Depth-One Homomorphic Federated Learning of Quantum Neural Networks
arXiv cs.DC (Distributed)
arXiv:2609.30581v1 Announce Type: cross Abstract: Encrypted training relies on keeping server-side updates low-degree. This constraint traditionally excludes models whose weights inhabit a compact Lie group (notably variational quantum…
-
Werracle: Sub-Cent Intra-Block AI Reflex Oracles and Flash-Loan Circuit Breakers for EVM Smart Contracts
arXiv cs.DC (Distributed)
arXiv:2609.30719v1 Announce Type: cross Abstract: Contemporary on-chain artificial intelligence (AI) encounters an intractable Von Neumann memory and latency wall. Storing static floating-point neural weight matrices inside Ethereum…
-
Beyond the Last Truffula Tree: SustainAI - A Water-Aware, Closed-Loop Framework for Environmentally Accountable AI
arXiv cs.DC (Distributed)
arXiv:2609.30747v1 Announce Type: cross Abstract: As artificial intelligence (AI) becomes embedded in everyday life, its environmental footprint, particularly water consumption remains largely invisible. While energy and carbon impacts are…
-
ITS Fairy: Occlusion Assistance Selected Against a Recipient's Own Perception Reports
arXiv cs.DC (Distributed)
arXiv:2609.31429v1 Announce Type: cross Abstract: Cooperative perception can expose object state beyond a vehicle's onboard sensors, but sensing occlusion can still leave a local safety application without the objects its collision…
-
Bandwidth-Aware and Cost-Efficient Pipeline Parallel Scheduling in Geo-Distributed LLM Training
arXiv cs.DC (Distributed)
arXiv:2605.25375v2 Announce Type: replace Abstract: The rapid evolution of large language models (LLMs) has made geographically distributed training necessary due to GPU scarcity within a single cloud region. In such cross-region settings…
-
Track me if you can: Ephemeral coin tracing
arXiv cs.DC (Distributed)
arXiv:2608.09249v2 Announce Type: replace-cross Abstract: Privacy-preserving payment systems are well understood, yet concerns about their misuse for financial crime have led to only limited adoption in regulated settings such as central…
-
Federated Targeted Maximum Likelihood Estimation
arXiv cs.DC (Distributed)
arXiv:2609.30503v1 Announce Type: cross Abstract: The evidence behind a scientific or operational decision is often held by hospitals, banks, or registries that cannot pool individual observations. Cross-silo federated learning moves…
-
Bandwidth, Latency, and 400 Million Kilometers: The Case for Mars-Local Compute
arXiv cs.DC (Distributed)
arXiv:2609.31566v1 Announce Type: new Abstract: There have been recent proposals for human settlements on Mars in 2030s. Any human activity on Mars must be preceded by extensive robotic exploration. However, Mars exploration is…
-
Managing Iterative Hybrid Quantum-Classical Optimization as a First-Class Scientific Workflow
arXiv cs.DC (Distributed)
arXiv:2609.30532v1 Announce Type: new Abstract: Today's Quantum Processing Units (QPUs) are too small and too noisy to solve large combinatorial optimization problems directly, so practical hybrid solvers split a problem into pieces and…
-
EAServe: Encode-Aware Disaggregated Serving for Multimodal Large Language Models
arXiv cs.DC (Distributed)
arXiv:2609.31551v1 Announce Type: new Abstract: Disaggregating the two stages, Prefill and Decode, onto separate GPU pools is now a standard optimization for (text-only) LLM serving. However, multimodal LLMs (MLLMs), which add a third…
-
The KV Cache Is the New Memory Wall
arXiv cs.DC (Distributed)
arXiv:2609.30854v1 Announce Type: new Abstract: Autoregressive LLM inference at long context is bounded by memory bandwidth, not arithmetic throughput, and the binding resource shifts from model weights to the Key-Value (KV) cache as…
-
Weave: Fine-Grained Dynamic SM Scheduling in an MoE Megakernel for Compute-Communication Overlap
arXiv cs.DC (Distributed)
arXiv:2609.21483v3 Announce Type: replace Abstract: Mixture-of-Experts (MoE) inference under expert parallelism (EP) turns each MoE layer into a distributed computation with costly dispatch and combine communication. State-of-the-art…
-
Cross-Backend QIEO: Universal Runtime Portability across OpenMP5, CUDA, HIP, and Multi-Language Interfaces
arXiv cs.DC (Distributed)
arXiv:2609.30914v1 Announce Type: new Abstract: Quantum-inspired algorithms emulate quantum mechanical principles, such as, superposition, interference, and probabilistic amplitude evolution, on classical hardware by representing candidate…
-
Unifying In-Memory Data Analytics through Sparse Compilation
arXiv cs.DC (Distributed)
arXiv:2609.30497v1 Announce Type: new Abstract: As modern data analytics workloads become increasingly heterogeneous and hardware-intensive, achieving efficient multi-core performance across diverse applications remains an open challenge…
-
Mixed Choice in Asynchronous Multiparty Session Types
arXiv cs.DC (Distributed)
arXiv:2602.23927v2 Announce Type: replace Abstract: We present a multiparty session type (MST) framework with asynchronous mixed choice (MC). We propose a core construct for MC that allows transient inconsistencies in protocol state…
-
Deduplication-while-Training: A Resilient Paradigm for Privacy-Preserving Cross-Client Deduplication in Federated Learning
arXiv cs.DC (Distributed)
arXiv:2609.31262v1 Announce Type: cross Abstract: Cross-client duplicate data in large language model training corpora degrades the efficiency of federated learning (FL) while exacerbating model memorization and privacy risks…
-
DynBranch: Speculative Subgraph Reuse for Dynamic Agentic LLM Serving
arXiv cs.DC (Distributed)
arXiv:2609.31047v1 Announce Type: new Abstract: Agentic LLM workflows decide their execution paths at runtime. Downstream computation may be predictable, or may have run before, yet it cannot begin until the model or the user resolves the…
-
FRESHLATENT: Channel-Aware Latent Adaptation for Resource-Constrained Embodied VLM Perception
arXiv cs.DC (Distributed)
arXiv:2609.30629v1 Announce Type: cross Abstract: Mission-critical UAVs increasingly rely on split vision-language model (VLM) perception under tight onboard-resource and wireless-communication constraints. However, corruption of…
-
Ambulance: saving BFT through racing
arXiv cs.DC (Distributed)
arXiv:2606.25099v2 Announce Type: replace Abstract: Today's practical Byzantine Fault Tolerant (BFT) state machine replication deployments are vulnerable to slowdowns. The main culprit is timeouts. Aggressive timeouts spuriously trigger…
-
Communication-Aware Model Distributed Inference via Latent Representation Compression
arXiv cs.DC (Distributed)
arXiv:2609.30413v1 Announce Type: new Abstract: We study optimization of distributed model inference over resource-constrained edge resources. We propose a framework that optimizes the trade-off between model accuracy and communication…
-
WeEnv: The Environment for Agentic Reinforcement Learning at WeChat
arXiv cs.DC (Distributed)
arXiv:2609.30766v1 Announce Type: new Abstract: Agentic reinforcement learning (RL) differs from conventional RL in that every task executes inside a complex environment, e.g., a virtual machine or a container. We find that agentic RL pays…
-
Distributed Learning as a Service: The Developer's Perspective
arXiv cs.DC (Distributed)
arXiv:2609.31061v1 Announce Type: cross Abstract: Application developers of distributed learning services face challenges that a typical federated learning loop does not address. Specifically, the model updates can still leak private data…
-
Deterministic Regime Switching and Feasibility Inversion in Dynamic Tensor Rematerialization
arXiv cs.DC (Distributed)
arXiv:2609.31250v1 Announce Type: cross Abstract: We report fine-grained, deterministic instability in Dynamic Tensor Rematerialization (DTR), an online eviction policy for memory-constrained DNN training, measured on the reference DTR…
-
HADIS: A Hybrid Architecture for Query-Aware Diffusion Model Serving
arXiv cs.DC (Distributed)
arXiv:2509.00642v3 Announce Type: replace Abstract: Query-aware model serving routes queries through cascades of increasing cost to balance quality and throughput. Existing cascade systems force all queries through the lightweight stage…
-
Predictive Rolling-Horizon Optimization for Commitment-Aware Model-Parallel Inference under Spatio-Temporal Edge Dynamics
arXiv cs.DC (Distributed)
arXiv:2609.31018v1 Announce Type: new Abstract: Model-parallel inference over dynamic edge systems requires scheduling decisions that account for not only instantaneous resources but also future resource contention and reliable service…
-
Momentum-Guided Federated Split Distillation for Personalized Temporal Edge Intelligence
arXiv cs.DC (Distributed)
arXiv:2609.31159v1 Announce Type: cross Abstract: We propose a momentum-guided federated split distillation framework for personalized, efficient, and autonomous temporal edge intelligence. We introduce TeRR-SAtt, our novel temporal…
-
Depth-adaptive Inference of Looped Language Models via Continuous Depth Batching
arXiv cs.DC (Distributed)
arXiv:2608.09444v2 Announce Type: replace-cross Abstract: A main promise of looped language models is depth-adaptive inference. By looping a block of shared layers a variable number of times, the model can use less compute for "easy"…
-
A Safety-Bounded SDC-to-MCP Gateway for Medical AI Agents
arXiv cs.DC (Distributed)
arXiv:2609.31358v1 Announce Type: new Abstract: The Model Context Protocol (MCP) provides a common interface through which AI applications discover and use external resources and tools. It allows language-model agents to ground their…
-
Authority at Commit Time: Reject-and-Rerun Semantics for Governed Agentic Systems
arXiv cs.DC (Distributed)
arXiv:2609.31490v1 Announce Type: new Abstract: Enterprise agents can compute for seconds or hours from a snapshot of policy, facts, task state, models, tools, and verifiers that may change before their work reaches the world. Completion…
-
Job Class Thermal Intent Aware Liquid Cooling Allocation for AI Data Centers
arXiv cs.DC (Distributed)
arXiv:2609.30785v1 Announce Type: new Abstract: GPU-dense AI data centers need to run on liquid cooling as air simply cannot shed the heat at these power densities. Yet the cooling loops themselves are blind to what workloads are about to…
-
KCensus: Synthesizing Latency-Optimal Consensus Fast Paths (Extended Version)
arXiv cs.DC (Distributed)
arXiv:2609.31302v1 Announce Type: new Abstract: Strongly consistent geo-replication often relies on fast paths to reduce latency in the common case of no failures or contention. Existing fast-path schemes, however, are ad hoc and…
-
Adaptive Switching Between Leader-Based and Leaderless BFT Protocols
arXiv cs.DC (Distributed)
arXiv:2609.31388v1 Announce Type: new Abstract: Byzantine fault-tolerant (BFT) protocols are known for providing operational consistency and resilience in distributed systems. However, evolving network conditions, often driven by the…
-
Scepsy: Serving Agentic Workflows Using Aggregate LLM Pipelines
arXiv cs.DC (Distributed)
arXiv:2604.15186v2 Announce Type: replace Abstract: Agentic workflows carry out complex tasks by orchestrating multiple large language models (LLMs) and tools. Serving them at a target throughput with low latency is hard because they are…
-
DeepFedNAS: Efficient Hardware-Aware Architecture Adaptation for Heterogeneous IoT Federations via Pareto-Guided Supernet Training
arXiv cs.DC (Distributed)
arXiv:2601.15127v4 Announce Type: replace-cross Abstract: Deploying federated learning across heterogeneous IoT device fleets requires tailored neural network architectures for each device class, yet existing Federated Neural Architecture…
-
Proceedings 19th Interaction and Concurrency Experience
arXiv cs.DC (Distributed)
arXiv:2609.30353v1 Announce Type: new Abstract: This volume contains the proceedings of ICE'26, the 19th Interaction and Concurrency Experience, which was held in Urbino, Italy, as a satellite event of DisCoTec'26. The ICE workshop series…
-
Bigger than CRISPR? A guide to the latest genome editors
Nature Technology
-
Prime editing goes large
Nature Technology