Knowledge catalogue
Excited to partner with the SpaceX team to scale up Composer. A meaningful step on our path to build the best place to code with AI. SpaceXAI and @cursor_ai are now working closely together to create
arXiv:2604.17513v1 Announce Type: new Abstract: Simulation frameworks such as Isaac Sim have enabled scalable robot learning for locomotion and rigid-body manipulation; however, contact-rich simulatio
arXiv:2511.00868v2 Announce Type: replace Abstract: Large Language Model (LLM) serving is increasingly constrained by the growing size of the key-value (KV) cache, which scales with both context lengt
arXiv:2604.16590v1 Announce Type: new Abstract: Accurate weather and climate prediction relies on data assimilation (DA), which estimates the Earth system state by integrating observations with models
HF becoming the platform for agents (assisted by their humans) to use and build AI (rather than just leveraging APIs)! Introducing ml-intern, the agent that just automated the post-training team @hugg
Hi i'm dwarkesh! Grew up all over the US, now sf-based and always down to nerd out about AI, science & history :) a lil about me: 🟠 Host of the dwarkesh podcast 🟠 Studied at UT Austin 🟠 Just published
I tested @huggingface ml-intern, given the prompt 'Fine-tune a Segment Anything Model (SAM) on a useful medical dataset. Train the model, and provide a comprehensive tutorial in a Jupyter Notebook fil
I’ve been using ml-intern for a while, and it genuinely changed my workflow. It's super good at: - Model/Dataset discovery. - Post-Training setup iteration. - Data processing workflows. Huge shoutout
Karpathy's autoresearch repo started an impressive trend. Agents can now train AI models to build SoTA agentic systems. And to think this is just scratching the surface. Ultimately, it boils down to g
arXiv:2604.18117v1 Announce Type: new Abstract: Post-training quantization (PTQ) is essential for deploying large diffusion transformers on resource-constrained hardware, but aggressive 4-bit quantiza
arXiv:2604.18090v1 Announce Type: new Abstract: Functional magnetic composites capable of large deformation, load bearing, and multifunctional motion are essential for next-generation adaptive soft ro
arXiv:2510.08726v2 Announce Type: replace-cross Abstract: Operator fusion has become a key optimization for deep learning, which combines multiple deep learning operators to improve data reuse and red
arXiv:2509.18169v3 Announce Type: replace-cross Abstract: Tasks on complex systems require high-precision numerical computation to support decisions, but current large language models (LLMs) cannot in
arXiv:2604.16583v1 Announce Type: new Abstract: Edge deployment of large language models (LLMs) increasingly relies on libraries of lightweight LoRA adapters, yet GPU/DRAM can keep only a small reside
arXiv:2604.17290v1 Announce Type: new Abstract: LLMs are widely used for code generation and mathematical reasoning tasks where they are required to generate structured output. They either need to rea
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
SpaceXAI and @cursor_ai are now working closely together to create the world’s best coding and knowledge work AI. The combination of Cursor’s leading product and distribution to expert software engine
Marcus Mendes / 9to5Mac: Apple says Tim Cook, as executive chairman, “will assist with certain aspects of the company, including engaging with policymakers around the world” — Apple confirmed today th
arXiv:2604.15359v1 Announce Type: cross Abstract: Understanding communication behavior in modern system-on-chip (SoC) designs is critical for functional verification, performance analysis, and post-si
AI agents are transforming how work gets done across all industries, accelerating everything from content creation to decision-making. NVIDIA’s expanded strategic collaborations with Adobe and WPP are
Eunice Xu / South China Morning Post: Chinese PCB maker Victory Giant surges ~57% in its Hong Kong trading debut after raising ~2.6B in its IPO, the largest in the city so far this year — Nvidia suppl
arXiv:2604.15472v1 Announce Type: cross Abstract: This paper is dedicated to lossless data compression with probability estimation using neural networks. First, we propose a probability estimation arc
NVIDIA researchers discovered a vulnerability in AI coding assistants where malicious software dependencies can inject harmful instructions into AGENTS.md configuration files, allowing attackers to re
Manufacturing is at an inflection point. Across every major industrial economy, the pressure to do more with less — due to faster design cycles, leaner operations and strain on skilled labor pools — i
arXiv:2506.10315v3 Announce Type: replace Abstract: Learned optimizers have been an active research topic over the past decade, with increasing progress toward practical, general-purpose optimizers th
The Korea Herald: SK hynix says it has begun mass production of the 192GB SOCAMM2, a next-gen LPDDR5X low-power DRAM module designed particularly for Nvidia's Vera Rubin — SK hynix Inc. said Monday it
arXiv:2509.13397v3 Announce Type: replace-cross Abstract: Social scientists are now using large language models to create 'silicon samples': synthetic datasets intended to stand in for human responden
We are excited to have @baseten as a day 0 launch partner for Kimi K2.6! Their inference stack brings KV-aware routing, NVFP4 on Blackwell, multi-modal hierarchical caching, and prefill-decode disaggr
Yann LeCun was right the entire time. And generative AI might be a dead end. For the last three years, the entire industry has been obsessed with building bigger LLMs. Trillions of parameters. Billion
Abram Brown / The Information: A look at Dylan Patel's SemiAnalysis, an AI newsletter and research firm that expects $100M+ in 2026 revenue from subscriptions and AI supply chain research — For Dylan
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
Qianer Liu / The Information: Sources: Google is in talks with Marvell Technology to develop a memory processing unit that works alongside TPUs, and a new TPU for running AI models — Google is in talk
That didn't take long for a total U turn. First the Nvidia CEO boasted AI would take a vast number of jobs. But now that there's a huge backlash against AI, he cobbles together this cope - to try to p
Reuters: Chinese lidar maker Hesai, a primary supplier for Nvidia's ADAS, announces EXT lidar, calling it the industry's first to integrate spatial and color detection — Hesai (2525.HK), China's leadi
1/ Old Nvidia chips are getting MORE valuable with age, not less. That has never happened in computing history. Marc Andreessen (@pmarca) on @latentspacepod with Alessio Fanelli (@FanaHOVA) and @swyx
Zvi Mowshowitz / Don't Worry About the Vase: A deep dive into Dwarkesh Patel's interview with Jensen Huang, including Huang's takes on Nvidia's moat and chip sales to China, and reactions to the inter
NVIDIA released PhysicsNeMo, an open-source AI framework designed to speed up nuclear reactor design through physics-based simulations. The framework uses machine learning-based multiphysics emulators
arXiv:2604.15180v1 Announce Type: cross Abstract: Sparse attention has been proposed as a way to alleviate the quadratic cost of transformers, a central bottleneck in long-context training. A promisin
Quantum computing keeps gaining momentum even though practically it’s years away from wide commercialization, and World Quantum Day April 14 provided an excuse for a lot of announcements. Among them:
arXiv:2511.10909v2 Announce Type: replace-cross Abstract: Modern AI accelerators rely on matrix multiply-accumulate units (MMAUs), such as NVIDIA Tensor Cores and AMD Matrix Cores, to accelerate deep
arXiv:2603.05493v2 Announce Type: replace Abstract: Effective robot autonomy requires motion generation that is safe, feasible, and reactive. Current methods are fragmented: fast planners output physi
arXiv:2311.04799v2 Announce Type: replace Abstract: Pretraining language models is still a challenge for many researchers due to its substantial computational costs. As such, there is growing interest
arXiv:2604.14552v1 Announce Type: cross Abstract: Modern datacenters increasingly rely on low-power, single-slot inference accelerators to balance performance, energy efficiency, and rack density cons
As artificial intelligence scales, AI infrastructure is becoming the deciding factor in whether enterprise AI delivers real value or stalls. AI is shifting from experimentation to unified systems wher
Ernie Image Turbo is a text-to-image generation model that can run efficiently on consumer-grade hardware like RTX 30 series GPUs with limited VRAM by using INT8 quantization. The post discusses techn
This post likely explores the various relationship dynamics and social connections people develop in shared living and working spaces, with a speculative or humorous tone suggested by the 'Maybe lover
Bloomberg: Hangzhou-based Manycore shares rose 187% early in its Hong Kong debut after raising $156M in its IPO; it is pivoting to selling AI training data to robot makers — Manycore Tech Inc. built a
arXiv:2601.14724v3 Announce Type: replace-cross Abstract: Recent advancements in Multimodal Large Language Models (MLLMs) have demonstrated significant improvement in offline video understanding. Howe
arXiv:2603.27844v2 Announce Type: replace Abstract: Majority voting over multiple LLM attempts improves mathematical reasoning, but correlated errors limit the effective sample size. A natural fix is
arXiv:2604.14825v1 Announce Type: cross Abstract: We present Nautilus, a novel tensor compiler that moves toward fully automated math-to-kernel optimization. Nautilus compiles a high-level algebraic s
Oh look! Anthropic's entire 'we are delaying Mythos' narrative was marketing hogwash. Kudos to FT for confirming what was obvious. Anthropic simply doesn't have the compute. FT: 'Multiple people with
Q-Day—the moment when quantum computers can break widely used cryptography—may be approaching faster than expected due to recent algorithmic and hardware advances. Advances in algorithms and design ar
Bloomberg: Sources: Cursor is in advanced talks to raise about 2B co-led by a16z at a pre-money valuation of more than 50B, with Nvidia participating — Cursor, a leading artificial intelligence startu
Financial Times: Sources: Recursive Superintelligence, a four-month-old start-up developing self-teaching AI and founded by ex-DeepMind and OpenAI engineers, has raised 500M+ — Group founded by former
arXiv:2604.14332v1 Announce Type: new Abstract: Diffusion-model inference and overdamped Langevin dynamics are formally identical. A physical substrate that encodes the score function therefore equili
Running LLMs across Cloudflare’s network requires us to be smarter and more efficient about GPU memory bandwidth. That’s why we developed Unweight, a lossless inference-time compression system that ac
arXiv:2603.25749v2 Announce Type: replace-cross Abstract: Arc-fault circuit interrupters (AFCIs) are essential for mitigating fire hazards in residential photovoltaic (PV) systems, yet achieving relia
Completely agree with Jensen! Not exporting AI out of fear is classic case where the cure would be 100x worse than the disease. You slow down innovation, progress and US technology and economic leader
arXiv:2604.13543v1 Announce Type: cross Abstract: Long Short-Term Memory (LSTM) neural networks have penetrated healthcare applications where real-time requirements and edge computing capabilities are
arXiv:2604.13327v1 Announce Type: cross Abstract: Modern GPU workloads, especially large language model (LLM) inference, suffer from kernel launch overheads and coarse synchronization that limit inter