Search: “routing”

Pangu-ACE: Adaptive Cascaded Experts for Educational Response Generation on EduBench

arXiv:2604.14828v1 Announce Type: new Abstract: Educational assistants should spend more computation only when the task needs it. This paper rewrites our earlier draft around the system that was actua

17 Apr 2026

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Towards Faster Language Model Inference Using Mixture-of-Experts Flow Matching

DGX agent

arXiv:2604.15009v1 Announce Type: cross Abstract: Flow matching retains the generation quality of diffusion models while enabling substantially faster inference, making it a compelling paradigm for ge

researcharxiv-cs-lg

17 Apr 2026

Dental-TriageBench: Benchmarking Multimodal Reasoning for Hierarchical Dental Triage

arXiv:2604.13060v1 Announce Type: new Abstract: Dental triage is a safety-critical clinical routing task that requires integrating multimodal clinical information (e.g., patient complaints and radiogr

16 Apr 2026

Actuation space reduction to facilitate insightful shape matching in a novel reconfigurable tendon driven continuum manipulator

arXiv:2604.12792v1 Announce Type: new Abstract: In tendon driven continuum manipulators (TDCMs), reconfiguring the tendon routing enables tailored spatial deformation of the backbone. This work presen

researcharxiv-cs-ro

DeferredSeg: A Multi-Expert Deferral Framework for Trustworthy Medical Image Segmentation

arXiv:2604.12411v1 Announce Type: new Abstract: Segmentation models based on deep neural networks demonstrate strong generalization for medical image segmentation. However, they often exhibit overconf

Local-Splitter: A Measurement Study of Seven Tactics for Reducing Cloud LLM Token Usage on Coding-Agent Workloads

arXiv:2604.12301v1 Announce Type: cross Abstract: We present a systematic measurement study of seven tactics for reducing cloud LLM token usage when a small local model can act as a triage layer in fr

QMC-Net: Data-Aware Quantum Representations for Remote Sensing Image Classification

arXiv:2604.11817v1 Announce Type: cross Abstract: Hybrid quantum-classical models offer a promising route for learning from complex data; however, their application to multi-band remote sensing imager

TempoNet: Slack-Quantized Transformer-Guided Reinforcement Scheduler for Adaptive Deadline-Centric Real-Time Dispatchs

arXiv:2602.18109v2 Announce Type: replace Abstract: Real-time schedulers must reason about tight deadlines under strict compute budgets. We present TempoNet, a reinforcement learning scheduler that pa

local-aiarxiv-cs-lg

ACE-TA: An Agentic Teaching Assistant for Grounded Q&A, Quiz Generation, and Code Tutoring

arXiv:2604.09572v1 Announce Type: cross Abstract: We introduce ACE-TA, the Agentic Coding and Explanations Teaching Assistant framework, that autonomously routes conceptual queries drawn from programm

Hardware

CodeQuant: Unified Clustering and Quantization for Enhanced Outlier Smoothing in Low-Precision Mixture-of-Experts

arXiv:2604.10496v1 Announce Type: new Abstract: Outliers have emerged as a fundamental bottleneck in preserving accuracy for low-precision large models, particularly within Mixture-of-Experts (MoE) ar

hardwarearxiv-cs-lg

DoReMi: Bridging 3D Domains via Topology-Aware Domain-Representation Mixture of Experts

model-releasesarxiv-cs-cv

arXiv:2511.11232v2 Announce Type: replace Abstract: Constructing a unified 3D scene understanding model has long been hindered by the significant topological discrepancies across different sensor moda

EvoESAP: Non-Uniform Expert Pruning for Sparse MoE

arXiv:2603.06003v2 Announce Type: replace Abstract: Sparse Mixture-of-Experts (SMoE) language models achieve strong capability at low per-token compute, yet deployment remains constrained by memory fo

MCERF: Advancing Multimodal LLM Evaluation of Engineering Documentation with Enhanced Retrieval

arXiv:2604.09552v1 Announce Type: cross Abstract: Engineering rulebooks and technical standards contain multimodal information like dense text, tables, and illustrations that are challenging for retri

Performance Characterization of Frequency-Selective Wireless Power Transfer Toward Scalable Untethered Magnetic Actuation

arXiv:2604.11645v1 Announce Type: cross Abstract: Frequency-selective wireless power transfer provides a feasible route to enable independent actuation and control of multiple untethered robots in a c

researcharxiv-cs-ro

Pyramid MoA: A Probabilistic Framework for Cost-Optimized Anytime Inference

arXiv:2602.19509v3 Announce Type: replace-cross Abstract: We observe that LLM cascading and routing implicitly solves an anytime computation problem -- a class of algorithms, well-studied in classical

Hardware

StreamServe: Adaptive Speculative Flows for Low-Latency Disaggregated LLM Serving

arXiv:2604.09562v1 Announce Type: cross Abstract: Efficient LLM serving must balance throughput and latency across diverse, bursty workloads. We introduce StreamServe, a disaggregated prefill decode s

hardwarearxiv-cs-ai

Dynamic sparsity in tree-structured feed-forward layers at scale

arXiv:2604.08565v1 Announce Type: cross Abstract: At typical context lengths, the feed-forward MLP block accounts for a large share of a transformer's compute budget, motivating sparse alternatives to

13 Apr 2026

Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection

arXiv:2604.08800v1 Announce Type: cross Abstract: Stepping-stone intrusions (SSIs) are a prevalent network evasion technique in which attackers route sessions through chains of compromised intermediat

13 Apr 2026

Alloc-MoE: Budget-Aware Expert Activation Allocation for Efficient Mixture-of-Experts Inference

arXiv:2604.08133v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) has become a dominant architecture for scaling large language models due to their sparse activation mechanism. However, the s

Efficient Quantization of Mixture-of-Experts with Theoretical Generalization Guarantees

arXiv:2604.06515v1 Announce Type: cross Abstract: Sparse Mixture-of-Experts (MoE) allows scaling of language and vision models efficiently by activating only a small subset of experts per input. While

Fail2Drive: Benchmarking Closed-Loop Driving Generalization

model-releasesarxiv-cs-cv

arXiv:2604.08535v1 Announce Type: cross Abstract: Generalization under distribution shift remains a central bottleneck for closed-loop autonomous driving. Although simulators like CARLA enable safe an

HQF-Net: A Hybrid Quantum-Classical Multi-Scale Fusion Network for Remote Sensing Image Segmentation

arXiv:2604.06715v1 Announce Type: cross Abstract: Remote sensing semantic segmentation requires models that can jointly capture fine spatial details and high-level semantic context across complex scen

Hardware

NestPipe: Large-Scale Recommendation Training on 1,500+ Accelerators via Nested Pipelining

arXiv:2604.06956v1 Announce Type: cross Abstract: Modern recommendation models have increased to trillions of parameters. As cluster scales expand to O(1k), distributed training bottlenecks shift from

hardwarearxiv-cs-lg

Selective Neuron Amplification for Training-Free Task Enhancement

arXiv:2604.07098v1 Announce Type: new Abstract: Large language models often fail on tasks they seem to already understand. In our experiments, this appears to be less about missing knowledge and more

A Responsibility-Centered Coordination Mechanism for Decentralized Multi-Agent Systems

arXiv:2610.21243v1 Announce Type: cross Abstract: Centralized orchestration provides comprehensive multi-agent coordination but creates bottlenecks in scalability, adaptability, and collective perform

ABNAT: Attention-Based Noise-Aware Training for Robust Transformers on Analog Compute-in-Memory Systems

arXiv:2610.22065v1 Announce Type: new Abstract: Transformer-based models have become the backbone of a wide range of tasks, but their intensive computation and data movement incur substantial energy o

Beyond the Dominant Label: Recovering Coexisting Lesion Evidence with Multiple Instance Learning

arXiv:2610.22091v1 Announce Type: new Abstract: Whole-slide images may contain coexisting lesions, yet weak supervision typically provides only the most severe diagnosis. We formulate this setting as

BiShield-TEE: On the (In-)Security of Unilateral Weight Obfuscation in On-Device TEE-Shielded LLM Partition

arXiv:2610.21565v1 Announce Type: cross Abstract: On-device deployment of large language models (LLMs) is increasingly adopted due to latency and privacy requirements, but exposes proprietary models t

BridgeAfford: Bridging Query-Target Gap for Generalizable 3D Affordance Grounding

arXiv:2610.21918v1 Announce Type: cross Abstract: Language-guided 3D affordance grounding aims to localize functional regions on 3D objects according to interaction intents. However, existing methods

CascadeLedger: Counterfactual Auditing of Information and Resource Integrity in LLM Cascades

arXiv:2610.20668v1 Announce Type: cross Abstract: We introduce CascadeLedger, a counterfactual execution-audit framework for budgeted language-model cascades. The protocol enforces call-before-read se

safetyarxiv-cs-cl

Compatibility Matters: A Framework for Token-Level Teacher Selection in On-Policy Distillation

arXiv:2610.21234v1 Announce Type: new Abstract: On-policy distillation (OPD) has recently been introduced as a promising paradigm for continual learning, where a student policy can be supervised at th

safetyarxiv-cs-lg

Condition-Contrastive Extrapolation for On-Policy Distillation of Language Models

arXiv:2610.21195v1 Announce Type: new Abstract: On-policy distillation (OPD) improves a student language model by training on its own generated trajectories with supervision from a stronger teacher. H

safetyarxiv-cs-ai

Explaining the Roles of Experts in Mixture-of-Experts LLMs for Personalization

arXiv:2610.22127v1 Announce Type: cross Abstract: Personalization of Large Language Models (LLMs) is increasingly important for user-centric applications, yet prior methods mainly adapt dense LLMs, wh

GraphBEAT: Graph-based Bucketized Event Aggregation Over Time for Early Risk Stratification of Gastrointestinal Cancers

arXiv:2610.22014v1 Announce Type: cross Abstract: Longitudinal electronic health records (EHRs) contain complementary diagnostic and laboratory information about patient health. However, irregularity,

Imperfect Is Perfect: Learned Column Prefilling for Branch-Price-and-Cut

model-releasesarxiv-cs-lg

arXiv:2610.21001v1 Announce Type: new Abstract: Column generation folklore warns that initial columns hurt: a warm start may bias the dual and slow the root of branch-price-and-cut. We show the opposi

Jev-LiteMAS: Separating Decisions from Generation in Multi-Agent Workflows

arXiv:2610.21411v1 Announce Type: new Abstract: Multi-agent LLM systems rely on generative models for expert execution, workflow planning, and artifact review. Yet many coordination operations require

Learning Not To Interfere: Adapting Experts For Model Merging

arXiv:2603.13467v2 Announce Type: replace-cross Abstract: Model merging has shown that multitask models can be created by directly combining the parameters of different models that are each specialize

Multi-SFT: Permissively-Licensed Multilingual Instruction-Following

arXiv:2610.22086v1 Announce Type: cross Abstract: Supervised fine-tuning datasets remain overwhelmingly English-centric and commercially restricted, as their content is frequently generated by proprie

safetyarxiv-cs-ai

Training-Free On-Policy Distillation

arXiv:2610.22236v1 Announce Type: cross Abstract: On-policy distillation consolidates specialist models into one student at the price of rollouts, teacher passes and gradient steps for every teacher.

safetyarxiv-cs-ai

Volition Elicitation: Operational Semantics for People and Their Machines

arXiv:2607.14138v2 Announce Type: replace-cross Abstract: The most prevalent distributed systems today include people and their personal machines (smartphones), and computations in them are driven by

Adaptive Observation Control for Active Visual Recognition

model-releasesarxiv-cs-cv

arXiv:2610.19624v1 Announce Type: new Abstract: Visual recognition models usually process a fixed view under a fixed computation budget, although images often benefit from different scales and amounts

Agent Sociodrama: Social-Dynamics-Guided Persona Recasting for LLM Agent Teams

arXiv:2610.15788v1 Announce Type: new Abstract: Large language model (LLM) agents are increasingly deployed as teams, yet they fail in characteristically social ways: information holders stay silent,

Cost-Efficient Privacy Enforcement for Email-Generating LLM Agents under Strategic Information Requests

arXiv:2610.18007v1 Announce Type: cross Abstract: LLM agents now act as delegated communication intermediaries: they read private context, draft messages, and send externally visible emails on behalf

Empowering Edge Agents: A Systematic Analysis of Edge-Cloud Collaboration in LLM Agent Execution

arXiv:2610.17816v1 Announce Type: new Abstract: Edge-cloud collaboration combines the lower inference overhead and greater data locality of edge-side execution with the stronger capabilities of cloud-

Energy Consumption Analysis of Discrete Diffusion and Autoregressive Language Models

model-releasesarxiv-cs-lg

arXiv:2610.19855v1 Announce Type: new Abstract: As Large Language Models (LLMs) continue to scale, their energy consumption during inference has become a critical concern. While Autoregressive (AR) mo

Enhancing Latent World Models with Associative Memory

arXiv:2610.19683v1 Announce Type: cross Abstract: World models trained via latent dynamics offer a scalable path for embodied planning, yet existing methods suffer from representational collapse, comp

Iterate on an LLM, Serve on an SLM

arXiv:2610.14576v1 Announce Type: new Abstract: As frontier LLMs gain more capabilities, the cost to run them also goes up. At the same time, small language models (SLMs) also gain more capabilities b

Language Steering for Multilingual In-Context Learning

arXiv:2602.02326v3 Announce Type: replace Abstract: If multilingual language model representations contain separable language-associated structure, then shifting between languages may require only sim