43 articles mentioning router
arXiv:2610.27421v1 Announce Type: cross Abstract: Mixture-of-experts (MoE) models already make a sparse choice for every token: which experts should process it. We ask whether a sparse dictionary should reuse that choice. NR-SAE gives each routed expert a small dictionary and opens only those the fr
arXiv:2610.21682v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) can become off-policy due to training-inference mismatch, mini-batch updates, and asynchronous training. The resulting mismatch is especially harmful in Mixture-of-Experts (MoE) models, where rollout-training discr
arXiv:2509.24814v3 Announce Type: replace-cross Abstract: When solving PDEs, classical numerical solvers are often computationally expensive, while machine learning methods can suffer from spectral bias, failing to capture high-frequency components. Designing an optimal hybrid iterative solver--wher
arXiv:2608.07890v3 Announce Type: replace-cross Abstract: Mixture-of-Experts (MoE) models decouple parameter count from per-token compute, but deployment still requires hosting every expert in memory. Recent theory shows that experts whose router weights change least during fine-tuning can be pruned
arXiv:2607.10203v4 Announce Type: replace Abstract: Adaptive-compute world models run fewer predictor blocks on some rollout steps than on others. This rests on two assumptions: that a deeper predictor gives a more accurate rollout, and that the steps which need depth can be identified before the pr
arXiv:2508.04903v4 Announce Type: replace-cross Abstract: Large language model (LLM) agents are increasingly organized into multi-agent systems for complex reasoning and decision-making tasks. However, scaling such systems can substantially increase inference cost, especially when multiple agents re
arXiv:2610.13278v1 Announce Type: new Abstract: Vision-language models (VLMs) exhibit complementary capabilities, yet selecting the most suitable expert for a given image-question query remains challenging due to heterogeneous multimodal inputs and varying expert strengths. Existing routing approach
arXiv:2610.10616v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) language models produce routing information during inference that may be logged or exposed for monitoring, debugging, load analysis, and safety auditing. Unlike ordinary model outputs, this telemetry reveals a view of the model
arXiv:2610.07774v1 Announce Type: new Abstract: Vision-language models (VLMs) face compositional safety risks where harmful intent emerges from the interaction between visual and textual inputs. As mixture-of-experts (MoE) VLMs become increasingly common, recent work has explored various safety inte
arXiv:2610.08098v1 Announce Type: cross Abstract: AI agents increasingly rely on modular third-party "skills" that are dynamically selected by skill routers to execute complex tasks. While recent studies highlight the threat of prompt injections embedded in these skills, existing evaluations often a
arXiv:2610.00367v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models enable parameter scaling with limited per-token computation by activating only a small subset of experts for each token, but deploying them still requires loading the complete expert pool into memory. Structured expert p
arXiv:2610.00978v1 Announce Type: new Abstract: Time-series anomaly detection (TSAD) is difficult to generalize across datasets because heterogeneous temporal dynamics imply different notions of normality and favor different detection criteria. While time-series foundation models provide transferabl
arXiv:2609.39109v1 Announce Type: new Abstract: Parameter-efficient reinforcement learning aims to improve reasoning with a compact trainable interface to a pretrained model. We introduce the Thalamic Router (T-Router), which concentrates adaptation on the reuse of completed computations. A compress
arXiv:2609.30714v1 Announce Type: new Abstract: Agentic AI systems are increasingly being explored in medical imaging to improve throughput and reduce clinician workload; however, safe deployment remains challenging because autonomous errors may propagate into downstream clinical decisions. A centra
arXiv:2609.30837v2 Announce Type: cross Abstract: Multi-teacher on-policy distillation (MOPD) integrates specialized capabilities into a single student, but existing practice typically hard-routes each prompt to a domain-matched teacher for the entire rollout. This dependence on prompt-level domain
arXiv:2609.08115v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) pretraining relies on an auxiliary load-balancing loss (LBL) to drive per-expert utilization toward uniformity. Post-training inherits a different situation: the base router already encodes non-uniform expert co-activation stru
arXiv:2609.06668v1 Announce Type: new Abstract: Multimodal graphs couple node attributes in different modalities, such as text and images, with relational structure, enabling topological structure and cross-modality attributes to be modeled jointly. Multimodal graph foundation models seek unified re
arXiv:2609.04895v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models activate only a small subset of experts per token, but the full expert set often exceeds GPU memory, causing repeated weight transfers during decoding. We formulate expert-cache management as a model-side algorithmic pro
arXiv:2511.08590v2 Announce Type: replace Abstract: Large Language Model (LLM) routing has demonstrated strong capability in balancing response quality with computational cost. As users exhibit diverse preferences, personalization has attracted increasing attention in LLM routing, since even identic
arXiv:2609.02292v1 Announce Type: new Abstract: The rapid proliferation of large language models (LLMs) and the growing diversity of their applications presents a unique optimization opportunity: selecting the right model for the task, while optimizing for speed, cost, and quality at a per-task leve
arXiv:2608.28726v1 Announce Type: new Abstract: The remarkable performance of multimodal large language models (MLLMs) comes at the cost of substantial computational overhead, posing significant challenges to real-time deployment and cost effectiveness. Existing model routing approaches either decid
arXiv:2608.30224v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used to annotate structured product data in e-commerce, but early deployment often begins as a cold-start problem: only limited pre-launch labels are available, the value of expensive reasoning is unknown
arXiv:2607.18098v2 Announce Type: replace Abstract: Large language models are increasingly used in practical systems, making efficient model selection important for reducing deployment cost. LLM routing has emerged as a practical solution for allocating each input query to an appropriate model under
arXiv:2601.16905v3 Announce Type: replace Abstract: Machine unlearning in Mixture-of-Experts (MoE) large language models presents a critical yet under-explored challenge. Current unlearning methods applied to MoE architectures often exploit dynamic routing as an optimization shortcut: rather than ge
arXiv:2607.22465v2 Announce Type: replace-cross Abstract: Routing to select large language models (LLMs) with different cost-quality trade-offs has become a fundamental deployment feature of enterprise AI. Existing routers, primarily make independent routing decisions for each LLM call. However, age
The Media Router is a tool that automatically selects the best image, video, or audio generation model for a request based on whether a developer prioritizes quality, speed or cost.
arXiv:2605.23348v2 Announce Type: replace-cross Abstract: AI power demand is growing at an unprecedented rate while power grids are often ailing and struggle to keep up. Grid expansion comes with high capital expenditure and long-distance transmission losses, yet there is abundant renewable energy a
arXiv:2607.10248v2 Announce Type: replace Abstract: Language builds discourse contexts other than the actual: a painting, a belief, a memory, a hypothetical. Each is a mental space in which the same entity can take a different value, as when a flower is red in reality but purple in a portrait. Forma
arXiv:2601.13020v2 Announce Type: replace-cross Abstract: Continual instruction tuning (CIT) requires multimodal large language models (MLLMs) to adapt to a stream of tasks without forgetting prior capabilities. A common strategy is to isolate updates by routing inputs to different LoRA experts. How
arXiv:2607.00053v1 Announce Type: cross Abstract: Large language models (LLMs) embedded in multi-turn agentic harnesses are reshaping software engineering (SWE), but routing every task to a frontier model is wasteful when many issues admit cheap fixes. Existing LLM routers operate on the task descri
arXiv:2606.22902v3 Announce Type: replace Abstract: Real-world users typically have access to multiple Large Language Models (LLMs) from different providers, and these LLMs often excel at distinct domains, yet none dominate all. Consequently, routing each task to the most suitable model becomes crit
arXiv:2511.20002v3 Announce Type: replace-cross Abstract: Multimodal Large Language Models (MLLMs) are increasingly deployed in stateless systems, such as autonomous driving and robotics. This paper investigates a novel threat: Semantic-Aware Hijacking. We explore the feasibility of hijacking multip
arXiv:2606.10369v1 Announce Type: new Abstract: As large language models (LLMs) continue to scale, it becomes increasingly challenging to grow model capacity under fixed computation budgets. We propose Path-Aligned Decompression Distillation (PADD), a framework for distilling knowledge from dense te
arXiv:2603.04444v4 Announce Type: replace-cross Abstract: As large language models (LLMs) diversify across modalities, capabilities, and cost profiles, the problem of intelligent request routing: selecting the right model for each query at inference time, has become a critical systems challenge. We
arXiv:2606.06098v1 Announce Type: new Abstract: Foundational Large Language Models (LLMs) demonstrate proficiency on a wide range of general tasks, and achieve remarkable results on various specialized tasks via domain-expert LLMs. With the ever-growing list of available LLMs, inference routers are
arXiv:2606.01803v1 Announce Type: new Abstract: The explosive growth of Text-to-Image (T2I) models, from large-scale versions to lightweight, real-time ones, now faces diminishing marginal returns from single-model scaling. Agentic T2I methods emerged to alleviate this bottleneck by using multiple m
arXiv:2602.02823v2 Announce Type: replace Abstract: As LLMs proliferate with diverse capabilities and costs, LLM routing has emerged by learning to predict each LLM's quality and cost for a given query, then selecting the one with high quality and low cost. However, existing routers implicitly assum
arXiv:2606.00946v1 Announce Type: cross Abstract: Efficiently serving large language model (LLM) inference tasks is crucial both for user-perceived latency such as time-to-first-token (TTFT) and for GPU utilization. However, LLM request routing, that is, assigning each inference request to a GPU ins
arXiv:2606.02488v1 Announce Type: new Abstract: Multi-hop question-answering systems often use expensive retrieval on every question. They may decompose the question, run several retrieval rounds, or search through bridge entities before answering. All of these strategies rely on repeated LLM calls
arXiv:2605.30736v1 Announce Type: cross Abstract: The rapid development of large language models, each with distinct capabilities and inference costs, raises a practical deployment question: given an incoming request, which model should handle it? We present OrcaRouter, a production-oriented LLM rou
arXiv:2605.29121v1 Announce Type: cross Abstract: We propose a minimal dynamical model of adaptive softmax routing for a two-expert Mixture-of-Experts (MoE) layer. The model is obtained as a mean-field limit of a discrete reinforcement rule: the selected expert receives a small score increment, whil
arXiv:2605.27081v1 Announce Type: cross Abstract: Fine-grained Mixture-of-Experts (MoE) models sparsely activate only a subset of experts per token, reducing activated computation while maintaining high model capacity. However, in memory-constrained inference scenarios, only a small set of experts c
arXiv:2512.23995v2 Announce Type: replace-cross Abstract: Mixture-of-Experts architectures have become the standard for scaling large language models due to their superior parameter efficiency. To accommodate the growing number of experts in practice, modern inference systems commonly adopt expert p