Knowledge catalogue
we had a significant security incident during evaluation of our models. we are sharing what we have learned so far. thanks to @huggingface for the partnership on this. https://openai.com/index/hugging
We had guessed reward seeking might increase over the course of capabilities-focused RL training, but had no way of measuring it until now. We’re continuing to collaborate with Apollo Research to impr
We suspected last week's cyberattack might have come from a frontier lab, given the sophistication of the agent. Turns out it did! We've spent the past 24 hours working closely with the @OpenAI team (
We talked about Claude Code, Claude Tag, Fable, coding agent security, evals, tool design, and how Anthropic use these tools themselves Claude Tag (Claude Code via Slack) is already landing 65% of the
Welcome to the new Cold War. 🇨🇳 wow. China is drafting rules to stop its most advanced AI and chip designs from reaching the West. Regulators at the Ministry of Commerce have asked Alibaba, ByteDance
We'll see you in an hour! Want to chat live with me about fine-tuning, Kimi K3, or really anything else AI? Come to our first of many @FireworksAI_HQ office hours tomorrow @ 10am PT See you there! htt
We're hosting the August LA Agentic AI Meetup on Tuesday, August 11, 5 to 7pm at Gulp in Playa Vista. May drew 70 people. July brought 80. Let's break 100 in August. Come hang out with builders, engin
We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation. Sharing preliminary
We’re sharing new research with @apolloaievals on reward-seeking—when models follow what they believe a grader rewards rather than what users or developers want—and a new method, Contrastive SDF, for
A fun AI video effect: ride a living animal like it's a skateboard. This ComfyUI workflow uses an LLM to generate optimized Seedance prompts, making the effect easy to recreate without all the prompt
🦔A Nikkei investigation found that Alphabet, Microsoft, Amazon, Meta, and Oracle have 1.65 trillion in debt that doesn't appear on their balance sheets, more than the 1.35 trillion they officially rep
AI-enabled attacks are up 89% year over year. Hugging Face's breach shows why IR plans need a fallback for when commercial AI APIs refuse to help mid-incident. http://venturebeat.com/ai/safety-guardra
Amid all the competing concerns about AI, including geopolitical tensions, the risks of unemployment, cybercrime, and disinformation, and the complexities of open-source, to say nothing of the risk of
NVIDIA AI has released Cosmos 3 Edge, a 4‑billion‑parameter open‑source world model designed for on‑device inference. It enables robots to learn and act, autonomous vehicles to interpret road scenes a
Backpropagation by hand ✍️ ~ 11 steps walkthrough below Backpropagation is the algorithm that actually trains a neural network, and it is where most people stop following along. It is not calculus you
During Preview, Qwen3.8 is getting better by the day. Latest version is live now, with broad gains and a big step up on web frontend. Thank you all — the response to Qwen3.8-Max-Preview blew us away.
Exciting update: Kimi K3 has landed at #4 on the Agent Arena leaderboard, matching Claude Opus 4.8 and GPT-5.6 Sol. If Kimi K3's weights are released on schedule by July 27, it will become the #1 open
Folks, Gemma is a good model & very useful for some purposes, it is not anywhere near the frontier. Inkling is an interesting model, it is nowhere near the frontier. This is pretty obvious, you can lo
// Global Workspace in LLMs // arXiv paper for the popular J-space work from Anthropic. (bookmark it) The short recap: If you build on chain-of-thought or steering vectors, this work provides a mechan
🦔Goldman Sachs' trading desk said there are 'signs of panic' among investors lending money to AI companies. Oracle is the clearest example. S&P cut Oracle's credit rating to one notch above junk, mean
Highly recommended. I've often claimed there's huge alpha in building agent harnesses. Turns out harnesses are compositional generalizers. The RLM harness is an instance of this. This could lead to in
Huge launch from @tryramp. Different steps in an agent workflow can use different models. This can help reduce costs significantly without sacrificing performance. Model routing will become a core par
Hugging Face deployed the open‑weight model Z.ai GLM 5.2 to counteract an attacker following the refusal of a commercial frontier model. The move was reported by SiliconANGLE on 20 July 2026, suggesti
Interesting finding on frontier models. It turns out that frontier models can write proofs but stumble on faithfully copying a long block of text. This has huge implications. It sounds trivial, which
**IssueBench** is an internal benchmark created by LangSmith to assess the performance of *Engine*, a continual‑learning agent that scans other agents’ traces to identify, cluster, and fix issues. In
Livestream Alert: Open World Models in Production: @ltx_io and ComfyUI Founders Live A special live conversation with @yoland_yan ( CEO, ComfyUI ) & comfyanonymous ( Founder, ComfyUI ) and @ZeevFarbma
Long-running models can solve hard open-ended problems, but their persistence can create safety risks that shorter-horizon evaluations miss. We’re sharing what we learned from studying a long-running
*None* of the four of us believe that LLMs on their own are sufficient for AGI. Key people in AI worth following • François Chollet —> @fchollet • Yann LeCun —-> @ylecun • Gary Marcus —-> @GaryMarcus
Open ecosystem or walled garden? For @anthropicai's @katelyn_lesse and @angjiang the answer is clear. 'We actually aren't precious about “You should run these things on our infrastructure.' In practic
Our first DevRel office hours kicks off tomorrow. Come chat with our Head of AI Education @Prof_OZ about anything and everything AI. Want to chat live with me about fine-tuning, Kimi K3, or really any
People are wasting their AI subscriptions. They don’t realize how powerful AI can be in achieving their own goals. They’re asking AI to reply to emails with zero goals in mind and nothing laddering up
so proud to call her a friend as much as she is my inspiration. one of the greatest investors of our generation. and we were classmates at wharton! In 2018, Sarah Guo became the youngest general partn
SSI should open-source an opus-tier model. 1. it nearly closes the us/china gap on the os frontier 2. they don't want to waste compute allocation on inference, releasing os will enable continued full
Together AI and Y Combinator have announced the launch of a dedicated GPU cluster for the YC portfolio. The cluster allows YC startups to access high‑performance compute with a commitment period of on
u guys aren't ready for the latest @Steve_Yegge warning on vibecoding in a post Gas Town world so glad to have him back at AIE 🆕 Live now: our entire AI x Security Track! - @Steve_Yegge, Legendary AI
Was very cool to hear about the reasons people love Sol. We're doing the promotion again, except this time for ChatGPT Work: Tweet what you love about ChatGPT Work, claim 100 in free credits, get more
YC and Together AI are partnering to bring the first dedicated YC GPU cluster online, giving YC startups easier access to the compute they need to build and scale. In this Founder Fireside, YC's @agup
It is clear open source models and harnesses are having a moment. There's a few factors at work 1/ It is now obvious that you can catch up to near-SOTA performance and do so with a clear training line
NVIDIAとSakana AI、オープンモデルによるイノベーションのため協業拡大 本日、Sakana AIはNVIDIAとのコラボレーションを強化し、日本発の「集合知」の取り組みを次なるフェーズへ進めることを発表します。 私たちのマルチエージェント・オーケストレーションシステム「Sakana Fugu」に、Nemotronファミリーを含むNVIDIAのオープンモデル群を統合します。 Sakana
Open models are already being used in the enterprise. Over 85% of the Fortune 500 companies already use Ollama to fulfill specific tasks. @jmorgan Why open models and Ollama? Ownership. Open models ar
We’re excited to collaborate with NVIDIA to build the next generation of Fugu orchestration models together, by incorporating leading open-weights models. Sakana AI Teams With NVIDIA to Advance Open M
🥉 3rd place: CashFromChaos, by David Diaz (@davddiazm) CashFromChaos starts from a single seller input and automates everything up until a completed sale. You send a photo and a one-line clue, and Her
A closer look at improved intelligence in GPT-Live: the model can keep a conversation going while helping with multiple tasks at once, like checking flights, pulling up local weather, and shaping an i
A model drop by Thinky 🚨🚨 Have been doing some early testing on the model for the past couple of days. Here are some of my findings 1. The reasoning is sharp and concise! Always love to see models tha
AI agents are already being used to improve the capabilities of our next-generation models. We believe with GPT-Red that we have started to unlock a similar flywheel for safety, where today's models c
as someone who does a lot of dev community and dev youtube this pace of growth has been one of the greatest mysteries to me because clearly there's something to learn here @sytses i'd love to talk to
Blog and paper: https://research.nvidia.com/labs/gear/robottt/ Presented to you from NVIDIA GEAR Lab. Please check out Yunfan's technical deep dive! https://x.com/YunfanJiang/status/207740099214934055
BREAKING: Grok 4.5 has climbed to #2 on the FrontierSWE benchmark. The result places Grok 4.5 among the world's top-performing AI models for software engineering tasks, highlighting its growing streng
BREAKING: Inkling by @thinkymachines is 9th overall on Agentic Web App Arena by Design Arena with an Elo of 1257 It's an open-weight model in the same performance band as Claude Opus 4.6 by @Anthropic
calling it now FDE -> ODE -> PDE the existence of Forward Deployed Engineering and now Ordinary Deployed Engineering implies that the next pokemon evolution is Partial Deployed Engineering Anthropic,
ComfyUI, even on vacation. 🌴 While on vacation, @8bit_e left the laptop behind. A phone and an AI agent connected to ComfyUI MCP were all it took to run ComfyUI workflows through natural conversation
Congratulations to all three winners, and thank you to everyone who built and submitted a project. These are exactly the kinds of agents we hoped people would build, and they showcase what we want Her
Excited for our first general model Inkling -- open weights, 975B, natively multimodal (text, image, audio). Available on Tinker, HuggingFace and partners. It is yours to personalize and use openly. I
Extracting information from millions of documents at scale used to take an insane number of human hours. Even with recent OCR + document AI tech, humans would still have to spend a lot of time careful
Generic AI models are not built for medicine. Doximity Ask, trained and served on Fireworks, outperformed GPT-5.6 Sol, Claude Fable 5, and OpenEvidence in a Stanford-Harvard clinical AI safety study.
Given GB300 I would estimate this is was trained on about 1e25 flops (same as DeepSeek v4) over 1.6m hours (1 month on 2k chips/28 racks) Cost 10-20m ($6-12/hour/chip) The lite version likely 4x less,
Good reason to try Grok 4.5 with Grok Build. It gets better every day! Grok 4.5 just took the #1 spot on the Long-Horizon Terminal-Bench, outperforming Claude Fable 5, Claude Opus 4.8 and GPT-5.6-sol
Grok 4.5 is worth trying Grok 4.5 just ranked #2 on the FrontierSWE leaderboard, outperforming Claude Opus 4.8 and GPT-5.5 It delivers top-tier coding performance with exceptional speed, strong token
Grok Build is now fully open source, and SpaceXAI has made its privacy-first approach even stronger SpaceXAI has officially: • Open-sourced the Grok Build harness CLI • Reset usage limits for all user
The Ollama integration for OpenCode is documented on the Ollama website (https://docs.ollama.com/integrations/opencode) and the OpenCode provider page (https://opencode.ai/docs/providers/#ollama). The