Knowledge catalogue
arXiv:2604.06291v1 Announce Type: cross Abstract: Low-Rank Adaptation (LoRA) enables parameter-efficient fine-tuning of Large Language Models (LLMs), and recent Mixture-of-Experts (MoE) extensions fur
arXiv:2610.21682v1 Announce Type: cross Abstract: Reinforcement learning (RL) can become off-policy due to training-inference mismatch, mini-batch updates, and asynchronous training.The resulting mism
arXiv:2610.19478v1 Announce Type: new Abstract: Agricultural systems make decisions over heterogeneous entity types, textual judgments, and knowledge-question options. Generating a textual answer adds
arXiv:2610.15603v1 Announce Type: new Abstract: Materials synthesis planning (MSP) connects AI-driven materials discovery to experimental realization by proposing precursors, synthesis operations, and
arXiv:2610.19625v1 Announce Type: new Abstract: We propose DAGFormer, a Transformer language model whose cross-layer connections are planned by a jointly trained topology predictor. Beyond the residua
arXiv:2610.19038v1 Announce Type: new Abstract: We study online task assignment with reusable resources. This problem arises in peer reviewer assignment, skill-based call routing, rideshare matching,
arXiv:2610.19179v1 Announce Type: cross Abstract: We study packet routing in queuing systems where queues act as selfish, myopic agents that use no-regret learning algorithms to select servers with un
arXiv:2610.19257v1 Announce Type: new Abstract: Memory for long-horizon video generation must satisfy two competing requirements: it must remain bounded so generation can continue indefinitely, while
arXiv:2610.20247v1 Announce Type: new Abstract: Multilingual models share capacity across languages, so adding languages can hurt performance on existing ones. Mixture-of-experts models expand this ca
arXiv:2610.19095v1 Announce Type: new Abstract: Chatbot Arena ranks language models with the Bradley-Terry model, which gives each model one score and so assumes Luce's choice axiom. Choice theory pre
arXiv:2610.19051v1 Announce Type: cross Abstract: Coordinators that act on what they infer about teammates are usually certified by the accuracy of those inferences. We argue that accuracy is the wron
arXiv:2610.17926v1 Announce Type: new Abstract: Autoregressive video generation relies on memory retrieval to model long-range dependencies. This capability becomes particularly critical for multi-sho
arXiv:2610.14672v1 Announce Type: new Abstract: Training-free sparse attention is a practical route to efficient video diffusion transformer inference, yet existing methods typically fix the competiti
arXiv:2601.22318v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly accessed as remotely hosted services by edge and enterprise clients that cannot run frontier models lo
arXiv:2610.18407v1 Announce Type: new Abstract: Diffusion language models reconstruct masked tokens in parallel, offering an alternative to autoregressive generation. Yet their Transformer backbones t
arXiv:2610.18857v1 Announce Type: new Abstract: Pixel-space diffusion models provide an appealing alternative to latent-space diffusion models by directly modeling raw images without relying on a lear
arXiv:2610.14644v1 Announce Type: cross Abstract: Fine-grained mixture-of-experts (MoE) models commonly increase expert count by narrowing each expert, treating expert shape as a by-product of routing
arXiv:2610.13481v1 Announce Type: new Abstract: On-policy distillation (OPD) provides teacher supervision on student-generated trajectories, but a single trajectory realizes only one continuation at e
Coding agents have changed how fast teams ship, and more of the software they ship now includes agents. A production agent spans model calls, routing decisions, tools, and the infrastructure underneat
arXiv:2604.03672v2 Announce Type: replace-cross Abstract: Government agencies must register, classify and route every citizen appeal within statutory time limits, and much of this work is still done b
Editor’s note: This article is adapted from Thomas Kurian’s keynote address at Gemini at Work 2026. It includes a number of announcements: Gemini, a single, universal agent for work that answers your
arXiv:2610.07729v1 Announce Type: new Abstract: Visual tokens are a major source of inference cost in vision-language models, yet simple image downsampling remains a surprisingly strong compression ba
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 mi
arXiv:2610.07591v1 Announce Type: cross Abstract: State tracking requires an update at every input, but the depth a Transformer applies to each token is fixed regardless of sequence length. We introdu
arXiv:2610.08137v1 Announce Type: cross Abstract: AI systems create images and videos with image/video generation models or by writing code and graphics descriptions that are then rendered. These rout
arXiv:2610.07332v1 Announce Type: cross Abstract: Long-horizon LLM agents are frequently implemented using sparse mixture-of-experts (MoE) models, yet the co-design of agentic behavior and MoE structu
arXiv:2610.03769v1 Announce Type: new Abstract: Multi-agent LLM systems pair a sender with broad context and an executor with a limited local view. We study when a short message improves the executor'
arXiv:2610.05409v1 Announce Type: new Abstract: Action-conditioned time-series forecasting requires accounting for how future actions and exogenous forcings influence multiple targets through partiall
This is an update. I posted LocalMind here many moons ago from another account, when it was a Gemma chat in a tab. LocalMind is a static web page that runs models on your GPU through WebGPU. It has no
arXiv:2610.06588v1 Announce Type: new Abstract: Long-horizon camera-controlled video generation relies on persistent memory to maintain scene consistency. Existing systems follow two strategies to ach
US three-row SUV options distinguish Tesla's supervised route driving from rivals' hands-free assistance, with current six-seat options and child-seat fit limits.
arXiv:2610.00663v1 Announce Type: new Abstract: Backdoored large language models (LLMs) can behave normally on benign inputs while producing attacker-specified outputs under hidden triggers. Existing
arXiv:2610.00992v1 Announce Type: cross Abstract: Learning-based driving planners are usually trained and evaluated in open loop against logged trajectories. In closed loop, a trajectory with small di
arXiv:2610.01296v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) Diffusion Language Models (DLMs) offer flexible parallel decoding and increased model capacity, but their large number of exper
arXiv:2510.23576v2 Announce Type: replace-cross Abstract: Urban micromobility applications, such as delivery robots, demand reliable navigation across large-scale urban environments while following lo
arXiv:2609.39839v1 Announce Type: new Abstract: Class-Incremental Learning (CIL) aims to continuously learn new classes without forgetting previously acquired knowledge. Parameter-efficient fine-tunin
arXiv:2609.39075v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) systems need inexpensive ways to route generated answers: accept low-risk outputs, review uncertain ones, and res
arXiv:2609.38484v1 Announce Type: new Abstract: Retrosynthesis enables the discovery of viable synthetic routes to target molecules. It plays a central role in modern drug discovery and materials desi
arXiv:2609.37062v1 Announce Type: cross Abstract: Dynamic layer skipping reduces LLM computation by allowing each token to execute only a subset of the model's layers. However, existing skippers rely
arXiv:2609.32971v1 Announce Type: new Abstract: Accurate forecasting of river discharge and floods is very challenging. River dynamics are affected by storage, meteorological forcing, and flow propaga
arXiv:2609.33581v1 Announce Type: new Abstract: Aerial Vision-Language Navigation (AVLN) requires UAVs to maintain reliable instruction following over long trajectories in complex 3D environments. How
arXiv:2609.32158v1 Announce Type: new Abstract: Moving obstacles can block a previously clear route while a quadruped robot executes a motion command. We investigate whether predicting changing cleara
arXiv:2609.34077v1 Announce Type: new Abstract: Mixture-of-experts (MoE) language models often exceed the memory of a single GPU. Expert offloading keeps most experts in host memory and loads them on
arXiv:2609.34759v1 Announce Type: new Abstract: Recent vision-language models (VLMs) have advanced vision-and-language navigation (VLN), enabling models to predict navigation actions from visual obser
arXiv:2609.34968v1 Announce Type: new Abstract: Vision-language-action and world-action models are increasingly popular, yet remain bottlenecked by physical interaction data that is scarce, institutio
arXiv:2609.35481v1 Announce Type: cross Abstract: Dynamic routing creates severe load imbalance in large-scale expert-parallel Mixture-of-Experts (MoE) training, turning GPUs that host hot experts int
arXiv:2609.33253v1 Announce Type: new Abstract: We present VGGT-Diff, a geometry-routed multi-view diffusion model for sparse-view novel view synthesis. Existing novel view synthesis (NVS) methods fac
arXiv:2609.31050v1 Announce Type: new Abstract: Efficient video generation requires reducing the quadratic cost of self-attention over long spatio-temporal token sequences. Existing efficient-attentio
We built an inference engine for MoE models that don't fit in VRAM + RAM. Most of the model stays on the SSD, and experts are read as tokens need them. This started as a proof of concept, and poc work
This is one of the strongest use cases I've seen for System One models like Jev. Pay close attention if you build custom harnesses. Jev is now available as a packaged router. What is it? You send the
arXiv:2609.30017v1 Announce Type: cross Abstract: Many LLM inference problems, including model routing, prefix-cache management, prompt trimming, and test-time search, can be viewed as optimization ov
I run a personal always-on agent fleet — eight seats claiming work in parallel from a shared task board, routing across Anthropic, OpenAI, xAI, Google, and a local Ollama floor. This week the spend fo
arXiv:2609.28448v1 Announce Type: cross Abstract: We study the nonequilibrium dynamics of a minimal recurrent transformer with N normalized tokens, Q=K=I, and a negative value map V=-I. Similarity-bas
arXiv:2609.26061v1 Announce Type: cross Abstract: Mixture-of-experts (MoE) models improve capacity with moderate overhead by sparsely activating experts per token. However, deploying MoE across resour
arXiv:2609.22813v1 Announce Type: new Abstract: We present commonsense ranked search (CoRS), a novel path planner that turns an abstract instruction into a route that follows commonsense. While existi
arXiv:2609.23745v1 Announce Type: new Abstract: Autonomous multi-drone navigation requires fleets to service distributed objectives in cluttered environments under tight computational budgets. Efficie
arXiv:2609.24635v1 Announce Type: new Abstract: When a language model reads an operation such as 'Swap the contents of Box F and Box B', its forward pass writes keys and values for those tokens into t
arXiv:2609.20824v1 Announce Type: new Abstract: We explore whether entropy-based confidence signals can be leveraged to improve the accuracy of Small Language Models (SLMs) with fewer than 3 billion p
arXiv:2606.15157v2 Announce Type: replace-cross Abstract: KV cache eviction methods typically use a single retention-rule family throughout a model, making eviction-method identity a model-level desig
I've been running Qwen3.8-Flash-Next as my main local coding model from past few weeks. The file is 95.5 GiB and my Mac has 64 GB. It works because the routed experts stay on SSD and only get read whe