Knowledge catalogue
arXiv:2604.26039v1 Announce Type: cross Abstract: The optimal kernel configuration for Mixture-of-Experts (MoE) inference depends on both batch size and the expert routing distribution, yet production
arXiv:2510.19669v4 Announce Type: replace Abstract: Recent reasoning Large Language Models (LLMs) demonstrate remarkable problem-solving abilities but often generate long thinking traces whose utility
arXiv:2604.25578v1 Announce Type: new Abstract: We present Marco-MoE, a suite of fully open multilingual sparse Mixture-of-Experts (MoE) models. Marco-MoE features a highly sparse design in which only
arXiv:2602.20730v2 Announce Type: replace Abstract: We study efficiency as a first-class objective in Neural Combinatorial Optimization (NCO) and present ECO, an efficient learning framework that comb
Pay attention to this one, AI devs, especially if you're thinking about agentic commerce or any agent network where many agents share hosts. A correct route to a cold agent is still a failed request f
arXiv:2601.21048v2 Announce Type: replace Abstract: Unsupervised neural combinatorial optimization (NCO) enables learning powerful solvers without access to ground-truth solutions. Existing approaches
TRINITY is a large language model coordinator system that evolves and improves its approach to orchestrating multiple LLMs or reasoning processes. The system likely uses evolutionary or adaptive mecha
Google Cloud Next ‘26 took place this week in Las Vegas, and the energy was incredible as we welcomed over 32,000 leaders, developers, and partners to explore the Agentic Era with us. Across three key
arXiv:2604.21187v1 Announce Type: cross Abstract: Ramsey-good graphs are graphs that contain neither a clique of size s nor an independent set of size t. We study doubly saturated Ramsey-good graphs,
arXiv:2604.21511v1 Announce Type: cross Abstract: Learned Sparse IR models, such as SPLADE, offer an excellent efficiency-effectiveness tradeoff. However, they rely on the underlying backbone vocabula
arXiv:2604.21525v1 Announce Type: new Abstract: Span-level skill extraction from job advertisements underpins candidate-job matching and labor-market analytics, yet generative large language models (L
arXiv:2604.19777v1 Announce Type: cross Abstract: Large Language Models (LLMs) exhibit a well-documented positional bias when processing long input contexts: information in the middle of a context win
arXiv:2509.21267v3 Announce Type: replace Abstract: Large language models often generate homogeneous outputs, but whether this is problematic depends on the specific task. For objective math tasks, re
arXiv:2603.11729v2 Announce Type: replace-cross Abstract: In recent years, RAPTOR based algorithms have been considered the state-of-the-art for path-finding with unlimited transfers without preproces
arXiv:2604.18788v1 Announce Type: new Abstract: Apple Neural Engine (ANE) is a dedicated neural processing unit (NPU) present in every Apple Silicon chip. Mixture-of-Experts (MoE) LLMs improve inferen
arXiv:2604.18759v1 Announce Type: new Abstract: Class imbalance is a widespread challenge in NLP tasks, significantly hindering robust performance across diverse domains and applications. We introduce
arXiv:2604.17709v1 Announce Type: new Abstract: Existing works on large language model (LLM) decomposition mainly focus on improving performance on downstream tasks, but they ignore the poor parallel
arXiv:2604.17425v1 Announce Type: new Abstract: Meta-optics promises compact, high-performance imaging and color routing. However, designing high-performance structures is a high-dimensional optimizat
arXiv:2604.16234v1 Announce Type: cross Abstract: Academic integrity continues to face the persistent challenge of examination cheating. Traditional invigilation relies on human observation, which is
In this post, we show you how to use Model Distillation, a model customization technique on Amazon Bedrock, to transfer routing intelligence from a large teacher model (Amazon Nova Premier) into a muc
arXiv:2604.12152v1 Announce Type: cross Abstract: Latent diffusion models for medical image super-resolution universally inherit variational autoencoders designed for natural photographs. We show that
arXiv:2604.12163v1 Announce Type: new Abstract: We present Nucleus-Image, a text-to-image generation model that establishes a new Pareto frontier in quality-versus-efficiency by matching or exceeding
Tesla's Full Self-Driving (FSD) version 14.3.1 successfully navigated a vehicle to five consecutive destinations without any human interventions, with routes inputted by Grok, Tesla's AI assistant. Th
arXiv:2602.01064v2 Announce Type: replace Abstract: Knowledge distillation has emerged as a pivotal technique for transferring knowledge from stronger large language models (LLMs) to smaller, more eff
arXiv:2604.09656v1 Announce Type: cross Abstract: Despite there now being more than 1,000 FDA-authorised AI medical devices, formal equity assessments -- whether model performance is uniform across pa
arXiv:2604.09702v1 Announce Type: cross Abstract: Precise segmentation of objects with highly similar shapes remains a challenging problem in dense prediction, especially in scenarios with ambiguous b
arXiv:2604.09414v1 Announce Type: cross Abstract: Learning-to-Defer routes each input to the expert that minimizes expected cost, but it assumes that the information available to every expert is fixed
arXiv:2604.06268v1 Announce Type: new Abstract: RL training of multi-turn LLM agents is inherently unstable, and reasoning quality directly determines task performance. Entropy is widely used to track
🏆 Qwen3.6-Plus just swept OpenRouter’s Daily, Weekly & Trending charts! The trial has wrapped. The model is fully live & production-ready. 📉 Lower latency | 🔥 Top-tier reasoning | 💰 Unmatched $/token
arXiv:2610.21055v1 Announce Type: new Abstract: Recursive self-improvement, in which AI systems build stronger AI systems, is one promising route to superintelligence. We study a weaker form that can
arXiv:2610.21361v1 Announce Type: new Abstract: Autonomous delivery in large scale buildings requires coordinated planning across multiple spatial levels, including building, floor, and room, while en
arXiv:2610.20634v1 Announce Type: new Abstract: Retrieval-augmented generation asks one question at volume: does this evidence support this claim? The answer is one bit, yet most pipelines obtain it f
arXiv:2610.22280v1 Announce Type: new Abstract: Computer-use agents increasingly tackle long-horizon tasks, where running a frontier model at every step is expensive and a small open-weight model alon
arXiv:2610.21685v1 Announce Type: cross Abstract: Adaptive assessment systems personalize learning by estimating learner proficiency and recommending suitable questions. However, many existing approac
arXiv:2610.20766v1 Announce Type: new Abstract: Imagined speech decoding from EEG offers a promising route toward naturalistic brain-computer interfaces. Traditional closed-set classifiers treat seman
arXiv:2610.22270v1 Announce Type: new Abstract: Surveillance cameras along urban roads produce large volumes of unlabeled video that could support the monitoring of public transport, but grouping the
arXiv:2610.18890v1 Announce Type: cross Abstract: Post-route power recovery changes cell choices under tightly coupled timing, electrical and physical constraints, and the value of each change depends
arXiv:2610.20263v1 Announce Type: new Abstract: Every multilingual speech pipeline begins with a language decision, an error there propagates silently through recognition, translation and routing. In
ggml-cpu : keep the FP16->FP32 lookup table on x86 with F16C (#30342) (#30343) ggml-cpu : keep the FP16->FP32 lookup table on x86 with F16C (#30342) #30270 routed both scalar conversions through F16C.
Cloudflare and OpenAI have formed a joint initiative to enhance the discoverability of AI‑generated content worldwide. By leveraging Cloudflare’s edge network, DDoS protection, and intelligent routing
arXiv:2610.14700v1 Announce Type: new Abstract: Most text-to-SQL benchmarks provide the target database. In a large data catalog, however, choosing the sources is part of the problem: an agent must fi
arXiv:2610.14625v1 Announce Type: new Abstract: Recursive self-improvement (RSI) offers a route to AI systems that participate in their own continued development. Recent methods allow language agents
arXiv:2610.20238v1 Announce Type: new Abstract: Graph Neural Networks (GNNs) excel on graph-structured data but suffer catastrophic forgetting when classes arrive sequentially. In class-incremental gr
arXiv:2610.20294v1 Announce Type: new Abstract: An agent can improve its behavior within an unresolved problem by changing how it acts and which experience it retains. We propose MMLA-RSI, a recursive
arXiv:2610.20174v1 Announce Type: cross Abstract: Reinforcement Learning on Mixed-domain preference data (Mix-RL) has become the mainstream approach for training general-purpose Generative Reward Mode
Meta’s Muse Spark Open is an 816‑billion‑parameter mixture‑of‑experts coding model that activates only 42 billion parameters per token through top‑8 routing across 256 experts, and has been trained on
arXiv:2610.18158v1 Announce Type: new Abstract: Learning dealer agents can settle on persistently wide quotes, raising what customers pay to trade. A venue can route more orders to tighter quotes, but
arXiv:2610.17456v1 Announce Type: cross Abstract: Language models are often evaluated alone but increasingly deployed together, routed between, made to debate, ensembled, or collaboratively trained. F
arXiv:2610.14982v1 Announce Type: new Abstract: Recurrent algorithmic reasoners are trained with per-step hints on small graphs and run for many more steps on large ones. On breadth-first search in SA
arXiv:2610.17552v1 Announce Type: new Abstract: Recent video super-resolution (VSR) models achieve efficient one-step generation through diffusion distillation. However, high-resolution inference rema
arXiv:2610.18142v1 Announce Type: new Abstract: Prompt optimization and generator post-training are two complementary ways to improve text-to-video (T2V) generation, and recent systems increasingly tr
arXiv:2610.17449v1 Announce Type: cross Abstract: Last-mile functional coverage closure for modern chip designs remains costly because a mature UVM environment exposes many controls while direct hits
arXiv:2610.14437v1 Announce Type: new Abstract: A platform that post-trains one base model for many business lines often has more lines than experts it can maintain, so lines must share experts, yet l
arXiv:2610.15830v1 Announce Type: new Abstract: When several predictions use the same internal representation, information about some targets can remain easier to recover than information about others
arXiv:2610.13239v2 Announce Type: replace Abstract: Visual grounding combines image--language understanding with geometric prediction, yet autoregressive multimodal models often express a box as a coo
arXiv:2610.18664v1 Announce Type: new Abstract: A patient's pathology case spans tens of slides from distinct specimens and anatomical sites, each carrying non-redundant, site-specific evidence. Formi
arXiv:2610.16082v1 Announce Type: new Abstract: Heterogeneous medical image data motivate domain-specialized diffusion experts, but expert specialization alone does not determine how their predictions
arXiv:2610.14682v1 Announce Type: new Abstract: Vision-and-language navigation (VLN) agents are typically evaluated by whether they reach a target and how efficiently they follow an instructed route,
arXiv:2610.14634v1 Announce Type: new Abstract: Continuous indoor-outdoor navigation connects destination-level reasoning with motion decisions that must respond to current observations. We present Vb
AWS engineers share kernel optimizations for fused Mixture-of-Experts (MoE) models on Amazon SageMaker AI: Swap-AB for WGMMA utilization, Triton backend routing for FP8 quantization, and eliminating T