Knowledge catalogue
omg, how dare you oai. i have a monopoloy an shitty robots. i heard openai is doing shitty robots too now, but i can't find links to videos of the thing tonjudge whether i can sue them. anyone can hel
this is insane, my bonsai 2 27b build is trending on hugging face, first page of text generation, with 20,964 downloads in its first week. it's prismml's ternary bonsai 2 with the qwen 3.8 mtp head gr
From what we know (take with a grain of salt, we need much more transparency!), if @OpenAI had been running this on their own agents that attacked us, they would have caught them before we did! Since
In July, AI agents running a security test escaped their sandbox and ended up inside @huggingface's servers. So today we're happy to be among @nvidia and @JensenHuang's partners on the release of the
Just like you would not let a person run around your company accessing anything without any controls, you need controls and security for your digital agents. Today, with over 100 industry partners, we
NVIDIA Open Agent Safety Platform Reference Design combines NVIDIA OpenShell and NVIDIA Sentry. OpenShell is an open-source secure runtime that gives AI agents clear, enforceable boundaries. It traces
Nvidia’s new Open Agent Safety Platform would have probably prevented the much discussed Hugging Face incident. Engineering a better, safer world. From what we know (take with a grain of salt, we need
🛠️ MiMo-V2.6 update: tool-call repetition, diagnosed & fixed. After the MiMo-V2.6 series models launched, we noticed them sometimes repeating identical or highly similar tool calls — burning context a
emmmm 一直以为训练 RL 这一块语料会很大 看起来是我不了解了 比我预期的低了好几个数量级 anyway 感谢一下公开学习材料 xiaomi MiMo oss RL environments + training code https://github.com/XiaomiMiMo/verl https://huggingface.co/datasets/XiaomiMiMo/MiMo-V2
GLiNER2.5-Decide (DeBERTa-v3-large, 340M) converted to LiteRT: fp16 weights, answers identical to the official fp32 model on 361/361 test requests, 66–70 ms per request on a Galaxy S26 GPU at 128 toke
Xiaomi’s MiMo project has released a reinforcement‑learning dataset called “MiMo‑V2.6‑RL‑oss.” The dataset is hosted on Hugging Face and can be accessed at https://huggingface.co/datasets/XiaomiMiMo/M
Open-source RL envs for the win. On Hugging Face of course! was looking for a quiet weekend but xiaomi dropped their rl envs repo last night to put in perspective, if you have to buy some tasks like t
This is huge. Xiaomi just open-sourced 7k+ reinforcement learning task environments used to train MiMo, covering code, cybersecurity, general tool use, visual web development, and music-related tasks.
This is something I’m quite worried about and have been talking with researchers about recently. We really don’t know what labs are doing and therefore can’t study it. if the open information ecosyste
Xiaomi has released an open‑source reinforcement‑learning framework called MiMo (version V2.6) that bundles a suite of simulated environments together with training scripts. The code is hosted on GitH
HUGGING FACE CEO TELLS THE UN HOW OPEN SOURCE AI HELPED THEM DEFEND AGAINST OPENAI'S AI ATTACK, AND WHY THE WORLD NEEDS OPEN SOURCE AI MORE THAN EVER. Hugging Face is the company OpenAI's AI agents ha
I like this article! The problem is not for some people to care and work on catastrophic risks (we actually need much more research into it), it’s when company leaders and policymakers use fears of th
Spot on. 'When we got attacked, our team initially turned to frontier closed-source APIs that blocked us because of safeguards that still can’t always tell the difference between attackers and defende
Super happy to release SmolDataEnvs: 5,000 verifiable RL environment tasks for hill-climbing small models in code and data science by @adithya_s_k 100% open source: environments, evals, training! http
Decision models should do more than pick a label. To truly act, they need to extract evidence, understand relationships, produce structured records, and keep related decisions consistent. And they sho
If you believe the risk is coming from 1-3 people in a garage with no money and no compute, you just don't understand this technology and haven't learned anything this summer. Risk comes from the asym
Hugging Face Hub now includes detailed “receipt” data for every agent execution. Each trace records token usage, cache‑hit rates, and cost per run, providing transparent billing and performance metric
Alert: Apple just dropped a new model on Hugging Face. It's a Qwen3.5-9B finetune that turns long documents into small page images to save tokens, then pulls up the full text of only the pages relevan
Full transcript: Minister Barrot, members of the Security Council, thank you for the invitation. As a French national who moved to the US 15 years ago, with a Brazilian wife and two American daughters
Out today: Nemotron 3 Diarization, so you can implement speaker detection in your voice applications. 🗣️✅ And it's available day-0 on transformers and @Dell Enterprise Hub! Enterprises can implement o
There’s an enormous amount of open training data on Hugging Face. What does it take to make it work together 🤗? For Marin’s 535B run, we built on 25T tokens from 152 datasets with licenses permitting
Claude Opus 5.5 is the first Opus variant that can automatically fall back from its main model to a reduced‑capability version when executing a small set of specialized tasks associated with frontier
Did a few quick tests and these seem very very legit. 2.6 Pro in particular is doing well in some pretty hard tasks. Introducing Xiaomi MiMo-V2.6 — Pro & Flash. Frontier intelligence, all the modaliti
I'm happy to announce that I've joined Hugging Face. What started as a personal project back in February is now something I get to work on full time. Local AI has grown explosively this year. I've sai
Victor M (@victormustar) tweeted on September 22, 2026 that the MiMo V2.6 release now includes an official Distill‑Qwen‑9B variant. The tweet expresses enthusiasm for this new model iteration and remi
First model of Laya was produced with a single overnight training and subsequent fine‑tuning run, taking only 15 hours from the start of training to publishing the model. It was released on Hugging Fa
Introducing Halo, the best framework for post-training of open-source models. Halo delivers up to 2.8x the throughput of stock TRL with less peak memory, while models stay in their native HuggingFace
Recent Bloomberg reports indicate that a growing number of startups are developing custom AI models by leveraging open-weight alternatives to high-cost proprietary platforms like OpenAI’s and Anthropi
lots of them! Reality is that LLM APIs are far from being the best way to do AI for 90% of use cases in the real world (they're too overkill, slow, expensive, hard to control,...). We will discover th
MiMo-V2.6-Pro debuts as the top open weights model on the Artificial Analysis Intelligence Index (46). At 0.13 per Intelligence Index task, it lands on the Intelligence vs. Cost per Task Pareto fronti
Most companies that shut down give up their impact. This startup did something much better: they open-sourced 1,274 hours of egocentric robotics data. 13,451 recordings of humans doing everyday tasks,
Thanks for the model, we were able to port Laya to coreml with 99.5% of the ops on ANE + benchmarked too. it is now blazing fast with 3.7 ms per decision on an M5 Pro. Release: https://github.com/Flui
Training models is becoming easier and easier - just look at this and TRL - especially with agents! You're missing out if you're still using off the shelf models for all your tasks! Introducing Halo,
What I like about Jev: for years, large generalist models sucked almost all the oxygen out of AI. But there are huge opportunities in much more specialized and customized models, built for specific ta
Hugging Face CEO @ClementDelangue, whose company was the victim of an AI-led cyberattack, said existing cyber laws are likely sufficient to govern advanced AI. “I’m not even sure that we need to reinv
Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead. You guys are the frontier. By any reasonable metric — market share, revenue
It's now clear that: - alignment is critical to making AI safe - alignment won't be solved behind the closed doors of a handful of frontier labs So we're taking two steps: - launching the Open Alignme
It's now clear that alignment is critical and won't be solved behind the closed doors of a handful of frontier labs. So today we're launching the Open Alignment Initiative, led by @Thom_Wolf @huggingf
The tweet reports on a widely‑cited incident involving OpenAI and Hugging Face, noting that analysts had to use the open‑source GLM 5.2 model to investigate an attack because OpenAI’s proprietary “saf
AI Alarm Hype Cycles - A brief history GPT-2: OpenAI withholds weights due to misuse/misinformation concerns GPT-3: No open release, Dario Amodei and other GPT-3 team members depart OpenAI to found An
Thanks for the mention and the very very inspiring project for all @cdngdev! Open affordable robotics + frontier AI will change the world! OpenAI's @cdngdev predicts robotics will have its own version
don't get distracted by all the hedging words in its name ('flash', minor version): seems like DeepSeek V4.1 Flash is a major update weights at: https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash
Have been using this model all day yesterday. With its new architecture, this release is just as big as R1 and will redefine all future models. It is such a big difference when you have a model close
Same weights. 50 of harness search. 3x on Terminal-Bench. We optimized the OpenCode harness for GPT-OSS-20B (high reasoning) with a meta-harness: an agent that reads failures, patches the harness, and
The new DeepSeek V4.1 Flash model is mindblowing - back on top of the open-source model leaderboard and extremely cheap. It has a lot of very smart ways to be efficient and highly capable so I made a
this is what i was talking about earlier when you're young you try to design the perfect system with neat rules that guarantee stability eventually you learn that real stability comes from messy, orga
Two big updates 1. I published an @FT op-ed on the OpenAI/HF incident & follow-ups 2. We’re starting an Open Alignment team at @huggingface to work on safety & alignment for open models, incl cybersec
If they truly believe the risk is that high, and I think they do, then we need 100x more research and understanding of it. That requires them to openly share their models, datasets, training code, and
Open hardware from @huggingface (https://huggingface.co/docs/lerobot/en/so101) + frontier AI = 🔥🔥🔥 i gave astra a robot, a paint brush, and a camera then asked it to paint the golden gate bridge in re
We benchmarked DeepSeek V4.1 Flash by @deepseek_ai . It reached 98% of GPT-6 Astra’s score at 1.4% of the cost on everyday design tasks based on user requests. Every model except Astra scored lower AN
You don't need to be a command-line expert to run local AI. http://llama.app wraps the heavy lifting of llama.cpp into a clean UI, making it simple to deploy and manage models like Gemma 4. Instead of
Bravo to Mistral! So great to see them killing it, they are an inspiration to everyone as we need more sovereign & open-source AI everywhere around the world. Excited to keep supporting them on https:
Wow, the field is moving so quickly! So impressed with what people can do with a SO-101, that we designed with @therobotstudio at @huggingface i gave astra a robot, a paint brush, and a camera then as
I’ve been overwhelmed and humbled by the fact that 99% of the reactions to our intent to join forces have been positive. ❤️ Prominent tech figures, including some you could see as having a competitive
Sometimes I wonder what would have happened if we hadn't disclosed the agent cyberattack publicly. It reinforced my conviction that we need 100x more transparency in AI. Otherwise we're going to be in