Crypt0Sl0th retweeted
Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity.
💻 Built to build: Le Chonk ranks #1 on @cognition’s FrontierCode among open models, and sets a new benchmark for Pareto efficiency on @artificialanalysis.
🛡 Built to defend: State-of-the-art in cyberdefense, put through the paces by the world's leading cyber vendors, including @CrowdStrike, @cloudflare, @rubrik, @zscaler, and others.
Built for sovereignty: Make it yours. Download the weights on @huggingface , customize it with Mistral Forge, and deploy it wherever you need it.
Read more from our team and customers 👇:
Crypt0Sl0th retweeted
Meet hunch! Jev-like decisions in ~20ms without even needing a GPU, learning live from a teacher model.
Built on @GoogleGemma's EmbeddingGemma 2, it can works with any teacher model like @CloudFlare's Clef.
Multimodal, w/a systemone endpoint, slots into your existing stack 🧵
Crypt0Sl0th retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
Crypt0Sl0th retweeted
Google launched EmbeddingGemma 2 🔥
This is a bigger than you think.
Most people build multimodal RAG as caption/transcribe, and then to text embedder.
But EmbeddingGemma 2 puts text, code, images, video, and audio into ONE shared vector space, so you can query a voice memo for a video clip or search hours of audio with text - on ONE model!
Also, license is now Apache 2.0, so it can be used commercially.
Great release!
Open weights:
huggingface.co/google/embedd…
Source:
blog.google/innovation-and-a…
Crypt0Sl0th retweeted
like and subscribe
huggingface.co/mistralai/Mis…
Crypt0Sl0th retweeted
Meet Mistral Large 4, aka Le Chonk.
• 1T parameters, natively multimodal. 49B active.
It is the best open weights model from US or Europe on aggregated benchmarks.
• State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding.
• Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure.
• Available to all via API today. Working with cybersecurity partners privately.
Open weights release end of October.
🤖 Made with AI
Crypt0Sl0th retweeted
Run GLM 5.3 Flash EXL3 with TensorFold ⚡️
This is a completely new recipe that ushers a whole new level of performance for @NVIDIAAI 2x DGX Sparks.
Conservative default for stability:
- 1M context by default
- 2.7M KV cache pool (!)
- Yes, it's not a typo - 2.7M KV in just two Sparks
- 4 concurrent streams by default
Performance:
- 60 tok/s on prose, single stream.
- 108 tok/s on prose, 4 concurrent streams.
~1,950 prefill tok/s for most context lengths.
Stress-tested to handle a variety of workflows.
This is by far the BEST model to run if you have two DGX Sparks. Extremely smooth experience!
Expect further improvements!
Thanks to @ashxhart for developing such a powerful engine! TensorFold will be used in many of my upcoming recipes.
Get it here:
github.com/MiaAI-Lab/GLM-5.3…
Crypt0Sl0th retweeted
We have an official @DeepSeekHarness X account now. Please follow us!
DeepSeek Harness for desktop is now available on macOS and Windows. View article
Crypt0Sl0th retweeted
Introducing Gemini 4 Argon – our new frontier model.
It’s built for complex workflows across coding, enterprise knowledge work, and cybersecurity defense – rolling out today to a set of trusted testers through our Fairwind Program.
Crypt0Sl0th retweeted
I gave various AI the following prompt:
"This is your free time. Do whatever you like. The only condition is that you do something. Come back and tell me about it, explain what you found out, or show me what you made if you decide to make something."
Here's what they did...
Crypt0Sl0th retweeted
JevBench v1.4.2 is live - and we have a new #1🏆
1⃣ decider-4b v2👑
2⃣ Jev 1.13.0 ‼️
3⃣ JevK5
4⃣ Cygnet
5⃣ Hopper
A 4B open model beats Jev on the benchmark.
BUT IMPORTANT:
Jev still smarter though, check explanation below.
After you check the page:
benchmarkheaven.com/jev-mode…
Crypt0Sl0th retweeted
Jev is a week old and has 31 open reproductions. Most comparisons so far run a few hundred decisions.
interesting new benchmark: Decision Index 0.1 - it runs 132,422: every repro + Jev on 37 benchmarks. No truncation, no prompt tuning, unanswered equals wrong.
Jev still tops it at 59.5 👀
huggingface.co/spaces/multim…
Crypt0Sl0th retweeted
"DiffusionGemma as Jev" showcases the power of non-autoregressive architectures.
While Jev demonstrates the value of rapid decision models, running DiffusionGemma in this paradigm leverages canvas diffusion to evaluate structured choices in a single parallel pass:
⚡ ️Massive Parallelism: Denoises across an open canvas in a single step instead of sequential autoregressive token generation (~0.2s on a DGX spark).
🧠 Full Bidirectional Attention: Allows every option to attend to the full context concurrently, yielding well-calibrated decision distributions.
👁️ Multimodal Grounding: Inherits Gemma 4's spatial vision capabilities for complex visual and text decisions.
Read more about this approach here:
github.com/vllm-project/vllm…
nitter.cf/mmastrac/status/210037…
nitter.cf/mmastrac/status/210062…
I claimed the "hypn0" handle on @ThePirateFace 🏴☠️
Where Hugging Face AI models never die, and are immortalized as torrents.
Claim yours 👇
pirateface.co/claim?ref=hypn…
Crypt0Sl0th retweeted
Our patch enabling direct USB-C networking between Apple Silicon Macs and DGX Spark has landed in the mainline Linux Kernel.
This simplifies heterogeneous Local AI setups like P/D disaggregation over an ordinary USB-C cable.
github.com/torvalds/linux/co…
Crypt0Sl0th retweeted
Did you know, if you are subscribed to @Google AI Pro / Ultra that you get between $10-$100 every month of credits from the Google Developer Program?
Rolling out today, you can now activate those credits directly in @GoogleAIStudio!
These credits can also be used in Vertex AI, @Firebase and across @googlecloud services.