Knowledge catalogue
arXiv:2605.01935v1 Announce Type: cross Abstract: Vision Mamba (ViM) models offer a compelling efficiency advantage over Transformers by leveraging the linear complexity of State Space Models (SSMs),
arXiv:2605.00174v1 Announce Type: cross Abstract: Video and image streaming on edge devices requires low latency. To address this, Neural Networks (NNs) are widely used, and prior work mainly focuses
arXiv:2605.00397v1 Announce Type: new Abstract: We present MiniVLA-Nav v1, a simulation dataset for Language-Conditioned Object Approach (LCOA) navigation: given a short natural-language instruction,
if even coding doesn’t make serious RoI for users, the jig is probably up. Economics of AI coding tools are crumbling. Cost per developer has doubled because cost of Claude code and microsoft copilot
The number of jobs in the future is endless because the problems to solve are endless. Jobs multiply as we get more complex. No AI or human can solve all problems and all the work to do in the Univers
The Information: Sources: Anthropic is in early talks to buy AI inference chips from UK-based Fractile when they become available in 2027 — As Anthropic's sales explode, straining the servers it uses,
arXiv:2604.27089v1 Announce Type: new Abstract: Large-language-models (LLMs) demonstrate enormous utility in long-context tasks which require processing prompts that consist of tens to hundreds of tho
arXiv:2509.16248v3 Announce Type: replace-cross Abstract: This paper presents GRAPHMEND, a high-level compiler technique that eliminates FX graph breaks in PyTorch 2 programs. Although PyTorch 2 intro
arXiv:2604.27128v1 Announce Type: cross Abstract: Foundation-model pipelines for individual-level livestock monitoring -- combining open-vocabulary detection, promptable video segmentation, and self-s
The U.S. Defense Department today announced that it has inked artificial intelligence procurement contracts with seven tech firms. The group includes Amazon Web Service Inc., Google LLC, Microsoft Cor
arXiv:2604.26334v1 Announce Type: cross Abstract: To usher in the next round of client AI innovation, there is an urgent need to enable efficient, lossless inference of high-accuracy large language mo
arXiv:2603.05811v2 Announce Type: replace Abstract: Current video generation models suffer from high computational latency, making real-time applications prohibitively costly. In this paper, we addres
arXiv:2604.25073v1 Announce Type: new Abstract: Deploying machine learning models under production constraints requires joint optimization over model family, quantization scheme, runtime backend, and
arXiv:2604.25421v1 Announce Type: new Abstract: Federated fine-tuning provides a practical route to adapt large language models (LLMs) on edge devices without centralizing private data, yet in mobile
arXiv:2604.24993v1 Announce Type: new Abstract: Randomized Smoothing (RS) offers formal ell_2 guarantees for arbitrary base classifiers but faces two key practical bottlenecks: (i) it often relies on
arXiv:2604.24809v1 Announce Type: new Abstract: We present Nautile-370M, a 371-million-parameter small language model designed for efficient reasoning under strict parameter and inference budgets. Nau
arXiv:2604.25884v1 Announce Type: cross Abstract: Quantum computing calibration depends on interpreting experimental data, and calibration plots provide the most universal human-readable representatio
arXiv:2604.24549v1 Announce Type: cross Abstract: Coordinating large populations of grid-edge devices requires learning methods that remain fully decentralised in deployment while still respecting thr
arXiv:2604.23632v1 Announce Type: new Abstract: Real-time text-driven joint audio-video avatar generation requires jointly synthesizing portrait video and speech with high fidelity and precise synchro
DigitalOcean describes their optimization and deployment of three large language models (DeepSeek V3.2, MiniMax-M2.5, and Qwen 3.5 397B) on their Serverless Inference platform, likely leveraging NVIDI
arXiv:2604.23858v1 Announce Type: new Abstract: Video generation, while capable of generating realistic videos, is computationally expensive and slow, prohibiting real-time applications. In this paper
arXiv:2604.24628v1 Announce Type: new Abstract: Proprietary design in commercial windrow-detection systems restricts transparency and limits progress in open autonomous forage-harvesting research. We
Welcome to the agentic era! Last week, leaders from our public sector customer and partner ecosystem took the stage at Google Cloud Next to share how they are leveraging AI and agents to scale their i
arXiv:2604.22062v1 Announce Type: new Abstract: There are 7,407 languages in the world. But, what about the languages that are not there in the world? Are humans so narrow minded that we don't care ab
arXiv:2604.22032v1 Announce Type: new Abstract: Every ML kernel ships with an implicit contract about what it computes. People rarely write the contract down. When two kernels disagree -- when a matmu
arXiv:2509.22166v4 Announce Type: replace-cross Abstract: The demand for efficient large language model (LLM) inference has intensified the focus on sparsification techniques. While semi-structured (N
The DeepSeek V4 garbled output bug in open source inference engine is fixed in SGLang. To everyone affected over the weekend, sorry for the trouble. Huge thanks to @Ant_Group for landing the fix PR. I
Dylan Patel announced his availability to meet in person at Dark Matter Coffee in New York City on a specific date between 11:00 AM and 12:30 PM. The post included a Google Maps link to the coffee sho
🎉 Day-0 support for @deepseek_ai V4 Pro and Flash on vLLM — a new generation of DeepSeek model, purpose-built for tasks up to 1M tokens. Alongside the release, we're publishing a first-principles walk
arXiv:2604.21026v1 Announce Type: new Abstract: Deploying large language models to heterogeneous hardware is often constrained by memory, not compute. We introduce MCAP (Monte Carlo Activation Profili
arXiv:2604.21076v1 Announce Type: cross Abstract: Medication reconciliation at clinical handoffs is a high-stakes, error-prone process. Large language models are increasingly proposed to assist with t
🔴 Trump vient de publier sur Truth Social un monologue de 1 500 mots sur le droit du sol, l'ACLU, les Indiens, les Chinois et la 'Savage Nation'. Ce texte inhabituellement long, inhabituellement cohér
arXiv:2604.19767v1 Announce Type: cross Abstract: We evaluate speculative decoding with EAGLE3 as an inference-time optimization for PayPal's Commerce Agent, powered by a fine-tuned llama3.1-nemotron-
arXiv:2407.09577v4 Announce Type: replace Abstract: Normalization layers are ubiquitous in large language models (LLMs) yet represent a compute bottleneck: on hardware with distinct vector and matrix
arXiv:2604.20046v1 Announce Type: new Abstract: 3D Gaussian Splatting (3DGS) has revolutionized novel view synthesis with high-quality rendering through continuous aggregations of millions of 3D Gauss
FLUX.2 [klein] 9B Base is a 9 billion parameter rectified flow transformer capable of generating images from text descriptions and supports multi-reference editing capabilities. The model fits in appr
arXiv:2409.07609v2 Announce Type: replace-cross Abstract: Deploying adversarially robust machine learning systems requires continuous trade-offs between robustness, cost, and latency. We present an au
arXiv:2604.19342v1 Announce Type: new Abstract: Generative AI-powered by Large Language Models (LLMs)-is increasingly deployed in industry across healthcare decision support, financial analytics, ente
Compute is king. I’m excited about the next phase of growth for Cursor, as their close partner. SpaceXAI and @cursor_ai are now working closely together to create the world’s best coding and knowledge
Last year at Google Cloud Next ‘25, we asked you to imagine a new future for AI. At Next ‘26, the question before you is how do you move AI into production across your entire enterprise? According to
arXiv:2604.18790v1 Announce Type: new Abstract: Depth completion from sparse LiDAR measurements and corresponding RGB images is a prerequisite for accurate 3D perception in robotic systems. Existing m
arXiv:2604.19398v1 Announce Type: new Abstract: Large language models (LLMs) are expensive to serve because model parameters, attention computation, and KV caches impose substantial memory and latency
Reuters: SK Hynix says it plans to invest ~12.85B in a new fab in South Korea for advanced packaging to meet AI demand for HBM, with construction starting this month — SK Hynix (000660.KS) said on Wed
arXiv:2604.06665v2 Announce Type: replace Abstract: Video depth estimation is essential for providing 3D scene structure in applications ranging from autonomous driving to mixed reality. Current end-t
arXiv:2604.18627v1 Announce Type: new Abstract: Ensuring human safety is of paramount importance in warehouse environments that feature mixed traffic of human workers and autonomous mobile robots (AMR
Google announced Ironwood, its seventh-generation TPU that is twice as power efficient as the previous generation, alongside specialized hardware designed to support the emerging agentic AI era. Ironw
This week at Google Cloud Next ‘26, we are sharing the evolution of Google Kubernetes Engine (GKE), delivering leading performance, efficiency, security, and scale for your most demanding and complex
While generative AI sparked a revolution, the true paradigm shift is the rapid evolution from standalone AI models to multi-agent autonomous systems. In this new era, the network transcends basic conn
arXiv:2604.16482v1 Announce Type: new Abstract: As vision-based robots navigate larger environments, their spatial memory grows without bound, eventually exhausting computational resources, particular
arXiv:2603.23224v2 Announce Type: replace Abstract: Generative models have shown substantial impact across multiple domains, their potential for scene synthesis remains underexplored in robotics. This
arXiv:2506.02264v3 Announce Type: replace Abstract: Building Task-Oriented Dialogue (TOD) systems that generalize across different tasks remains a challenging problem. Data-driven approaches often str
“Cursor has also given SpaceX the right to acquire Cursor later this year for 60 billion or pay 10 billion for our work together.” personally this is the most exciting option pricing deal of the year,
arXiv:2604.17789v1 Announce Type: cross Abstract: The MXFP4 microscaling format, which partitions tensors into blocks of 32 elements sharing an E8M0 scaling factor, has emerged as a promising substrat
SAUVER LA FRANCE ET L'EUROPE EN FAISANT LE SEUL PARI QUI VAILLE (série banane rouge) I. COMME BONAPARTE : CONCENTRER LES FORCES, PAS LES DILUER Dans le discours politico-industriel d'aujourd'hui, on e
arXiv:2604.17627v1 Announce Type: new Abstract: Serving large language models under latency service-level objectives (SLOs) is a configuration-heavy systems problem with an unusually failure-prone sea
arXiv:2604.15768v1 Announce Type: cross Abstract: AI-driven methods have demonstrated considerable success in tackling the central challenge of accurately solving the Schrodinger equation for complex
We’re launching Kimi K2.6 on Fireworks as a Day-0 launch partner! K2.5 was the base for standout models like @Cursor’s Composer 2 and was the most popular model on our training platform. K2.6 on Firew
What a week in SF for Gemma! 💎 - Gemma 4 SF Meetup, with Unsloth, MLX, VLLM, NousResearch, and Cactus Compute - Hackathon at Y Combinator with Cactus Compute - Ollama Gemma 4 Meetup, with SGLang - NVI
Emily Shugerman / The San Francisco Standard: A profile of Maria Davidson, who heads California Renewal, a pro-business political group backed by Silicon Valley power players, seeking to raise $100M i
Release b8833 of llama.cpp includes updates to the ggml-webgpu backend, fixing compiler warnings and refactoring FlashAttention encoding, along with workflow improvements and precision adjustments for