Knowledge catalogue
Yann LeCun's latest talk at ETH Zürich 'You should not work on LLM. At least if you're in academia, you should absolutely not work on LLMs. There is nothing you can bring to the table. If you're inter
Mishal Husain / Bloomberg: Q&A with Google SVP and DeepMind Institute co-director James Manyika on AI risks and why responsibility must be shared across industry, government, and society — The accord,
Bradley Olson / Axios: Sources: several Western open-weight models are set to launch this month, including Reflection AI's first model, which will rival top Chinese open-weight models — A closely watc
Meir Orbach / CTech: Nasdaq-listed neocloud Nebius acquires Inferize, whose tech helps optimize GPU utilization and reduce AI request costs, sources say for 100M to 150M — The Israeli startup, which w
Banger paper from NVIDIA on long running agents. A model can accept 128K tokens of context and still make more mistakes the longer it works through a task. If your agent loses its place partway throug
arXiv:2610.00321v1 Announce Type: cross Abstract: Speculative decoding accelerates large language model inference by drafting future tokens cheaply and verifying them with the target model in parallel
arXiv:2610.01439v1 Announce Type: new Abstract: Parallel drafting reduces the drafting overhead of speculative decoding for large language models (LLMs), but its gains remain limited by the accepted p
arXiv:2610.01967v1 Announce Type: new Abstract: As large language models (LLMs) keep growing in size and complexity, their training frameworks evolve at a rapid pace as well. Therefore, continuous int
I've been experimenting with running larger models locally with Ollama, mainly because I'm interested in using them for agentic coding. I kept seeing benchmarks expressed as something like: 40 tokens/
arXiv:2610.01445v1 Announce Type: new Abstract: UMAP achieves scalable layout optimization through stochastic negative sampling. However, this stochasticity can lead to unstable embeddings across reru
arXiv:2610.01468v1 Announce Type: cross Abstract: Low-cost NAS must both explore high-performing architectures and identify them reliably, yet reducing evaluation cost often weakens the fidelity of ca
Hi, i got QFN running on my single r9700 but im not sure if i did everything right to get the best quality and speed out of this setup. Dont want to annoy anybody, maybe someone with the same card can
arXiv:2602.12630v2 Announce Type: replace-cross Abstract: Most large language models (LLMs) run on external clouds: users send a prompt, pay for inference, and must trust that the remote GPU executes
The Comfy Dev Platform Challenge is open! Partners: @ltx_io & @bfl_ai For two weeks (10/5–10/19), we're challenging the community to open-source a paid AI feature. 🏆 $10K grand prize 🖥️ 3x NVIDIA RTX
arXiv:2610.01022v1 Announce Type: new Abstract: Memory-attention-based Video Instance Segmentation (VIS) methods have demonstrated strong zero-shot tracking capability, yet their substantial memory re
arXiv:2610.01959v1 Announce Type: cross Abstract: Path finding and multi-robot motion planning require trajectories that are smooth, goal-directed, and collision-free in environments with complex geom
The neocloud market is moving past its origins as a stopgap for scarce graphics processing units. AI-native startups now choose their infrastructure on latency, burst capacity and openness, not just c
arXiv:2609.38438v1 Announce Type: cross Abstract: Machine learning surrogate models offer a promising path toward accelerating plasma turbulence simulations. We present PreVAE-Turb, a surrogate modeli
arXiv:2609.39111v1 Announce Type: new Abstract: Human intelligence relies heavily on learned intuition: recognising patterns and judging situations without explicitly unfolding every intermediate step
Learn how to use Amazon S3 Vectors as the persistent memory layer within the NVIDIA NeMo Agent Toolkit (NAT), deployed on Amazon Elastic Kubernetes Service (Amazon EKS). This post shows how NAT's memo
arXiv:2609.38570v1 Announce Type: new Abstract: Class 8 trucks differ from passenger cars in geometry, dynamics, and maneuvering requirements. As a result, vision-language-action (VLA) models trained
arXiv:2609.38617v1 Announce Type: new Abstract: Legged robots can learn expressive whole-body skills from the motions of humans and animals. Due to the morphology gap between the source and the robot,
arXiv:2511.10333v2 Announce Type: replace-cross Abstract: Training large language models (LLMs) at scale incurs substantial communication overhead, while static gradient compression cannot adapt to gr
arXiv:2609.40093v1 Announce Type: cross Abstract: Expert parallelism (EP) enables inference of large Mixture-of-Experts (MoE) models by placing their experts across multiple GPUs, but requires substan
Mackenzie Hawkins / Bloomberg: Experts say mounting cases of Nvidia chips reaching Chinese AI companies despite US export controls are increasingly pointing to gaps in Nvidia's due diligence — Nvidia'
Spooky season is streaming in. Alongside falling leaves, pumpkin spice and everything nice, 25 new games are joining GeForce NOW throughout October, including six ready to play this week. From a new C
arXiv:2609.38198v1 Announce Type: new Abstract: Diffusion maps, and kernel methods more generally, provide an interpretable nonlinear spectral representation basis for geometric learning. In the geome
Generative AI has made remarkable progress in creating increasingly realistic representations of the world. Image models can synthesize photorealistic scenes, video models can generate complex dynamic
arXiv:2609.38312v1 Announce Type: cross Abstract: The Platonic Representation Hypothesis predicts that sufficiently scaled foundation models converge on a shared representation of the world. As each n
arXiv:2609.40297v1 Announce Type: new Abstract: Autonomous exploration demands that robots continuously evaluate candidate viewpoints based on their expected information gain and execution cost. Sampl
GPT-6 Astra Ultrafast, running on NVIDIA Blackwell GPUs, is available now in the OpenAI API and to eligible ChatGPT Work and Codex users. Accelerated by inference optimizations through OpenAI’s models
arXiv:2609.38900v1 Announce Type: new Abstract: Streaming video understanding requires models to process unbounded visual streams while preserving rich visual semantics across vast temporal horizons,
Storage architecture is being rewritten for artificial intelligence factories. Traditional enterprise storage was designed around workloads that scaled in relatively predictable ways. AI changes that
Artificial intelligence factory economics increasingly depend on more than access to high-performance graphics processing units. As agentic systems draw on multiple models, databases and tools, the en
arXiv:2609.38511v1 Announce Type: new Abstract: Autonomous underwater robots require robust perception, estimation and control to operate in confined environments. This paper presents an open-source B
AI factories are built by the megawatt, even by the gigawatt. Each megawatt factory costs roughly $60 million, and AI factory operators will only commit capital on that scale with a clear view of the
arXiv:2609.39223v1 Announce Type: new Abstract: Large language model (LLM) inference is increasingly moving toward lower precision to realize the throughput of hardware accelerators, but aggressive po
arXiv:2609.39082v1 Announce Type: new Abstract: As new evidence arrives, a sequence model must update what it remembers and how memory influences predictions. While Transformers incur computation and
The Information: SoftBank made the final 10B investment in its 30B pledge to OpenAI's last funding round; source: Nvidia also made its final 10B investment in the round — Nvidia and SoftBank have each
arXiv:2609.38830v1 Announce Type: new Abstract: Sparse attention is widely used to accelerate long-context inference in modern large language models (LLMs), but its input-dependent execution behavior
arXiv:2609.38984v1 Announce Type: cross Abstract: World-action models (WAMs) leverage pretrained video models to improve generalization in robot control by jointly predicting future visual states and
arXiv:2609.39704v1 Announce Type: cross Abstract: This paper demonstrate that whether masking-based token pruning helps or hurts worst-group robustness can be predicted before deployment, without labe
arXiv:2609.36800v1 Announce Type: new Abstract: Compiler backends are expensive to build and maintain as programming models, workloads, and accelerators evolve. We investigate whether large language m
arXiv:2609.36222v1 Announce Type: new Abstract: Large language models are increasingly expensive to serve. In large-scale serving systems, autoregressive decoding is often bottlenecked by transferring
arXiv:2609.36735v1 Announce Type: new Abstract: In branch-and-bound (B&B) for mixed-integer linear programming (MILP), branching variable selection critically impacts efficiency. Existing neural branc
Amazon Bedrock AgentCore Runtime Instances gives multi-agent workflows AWS managed EC2 infrastructure with GPUs, persistent volumes, and multi-day sessions. In this post, we deploy a three-agent music
Cognition is the first customer running @NVIDIA Vera Rubin, powered by @CoreWeave! Hopper in 2022. Blackwell in 2024. Now Vera Rubin: On SWE-2 inference, the new chips deliver about 4.8x more token th
Token economics are becoming a defining measure of artificial intelligence infrastructure efficiency. As AI workloads move beyond training, the economics of producing useful intelligence are starting
Generative recommender systems model recommendation as high‑cardinality event streams, turning user actions, context, and candidate items into tokens for sequence modeling over user behavior. NVIDIA’s
arXiv:2609.37829v1 Announce Type: new Abstract: Video diffusion transformers (DiTs) increasingly adopt mixture-of-experts (MoE) architectures to reduce active computation, but their full expert storag
NVIDIA has released its cuObject client and server libraries, providing RDMA‑based APIs that allow GPUs to access object storage directly without routing data through the host CPU. The xio‑sig consort
arXiv:2609.36757v1 Announce Type: new Abstract: Diffusion-based video restoration recovers realistic details, but its practical deployment is limited by two efficiency bottlenecks: costly VAE encoding
arXiv:2605.06206v2 Announce Type: replace Abstract: Mixture of experts has emerged as the primary mechanism for making Large Language Models (LLMs) computationally efficient. However, in distributed s
Building on nearly a decade of co-engineering, CoreWeave has built NVIDIA compute, networking and software into a cloud purpose-built for AI that’s still returning on investment across multiple genera
arXiv:2609.36074v1 Announce Type: new Abstract: Memory-efficient scaling on clustering problems without sacrificing statistical accuracy is of central interest for large-scale data analysis and machin
Phoebe Liu / The Information: GPU cloud provider GMI Cloud raised 668M, including 223M in equity led by ARCHIV with participation from Nvidia and 445M in credit led by Taiwanese bank CTBC — GMI Cloud,
arXiv:2609.38123v1 Announce Type: new Abstract: World simulation is inherently multisensory, demanding synchronized visual and acoustic dynamics in real time. Yet prevailing interactive world models r
I got thousands of DMs and it looks like @bot can't automatically analyze DMs (when @elonmusk?) so it will take me a few days to go through all of them, sorry! If you don't get an answer, don't take i
arXiv:2609.37481v1 Announce Type: new Abstract: Large-scale pervasive sensing increasingly relies on high-resolution satellite imagery, yet task-specific onboard vision is constrained by costly annota
Hi again! This is a update to my previous post MageTrail V0.25 where I continued this tech demo toy model thingy called MageTrail, a Danbooru/E621 proof of concept Full-Finetune of Microsoft's MageFlo