NASA Puts Google’s Gemma Large Language Model in Orbit
The viability of orbital data centers hosting the largest and most capable large language models (LLMs) remains hotly contested. But enormous deployments that require thousands of…
Model archive
AI models 28 tracked
Every primary-source story across every tracked model. Filter by clicking a chip.
RT Nat Gurlain: Backup of original transcript (Liang Wenfeng DeepSeek investors call)- https://archive.ph/NLuG9
RT Nat GurlainBackup of original transcript (Liang Wenfeng DeepSeek investors call)- https://archive.ph/NLuG9Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞): Full transcript of Wenfeng's investor conference call. He's a…
Minimax 2.7 vs DSv4 flash vs laguna S 2.1
I got 192 GB of Vram to use and a long horizon agentic task about coding, I would like opinion from ppl who have had extensive…
Google CEO Pichai says Gemini's next leap depends on building "much larger base models"
Alphabet has raised its 2026 investment forecast to as much as $205 billion, saying demand continues to outpace spending. Google Cloud grew 82 percent in the…
Experts say exploiting Anthropic’s Fable isn’t how Kimi K3 got so good
"I don't think you get a model this strong and this quickly on the heels of Fable doing strictly distillation," one expert told TechCrunch.
DeepSeek Founder’s 4-hour investor meeting: DeepSeek is prioritizing AGI over user growth and commercialisation
A Chinese article compiled 52 remarks from Liang Wenfeng’s four-hour investor meeting. I’ve summarised the most important ones below. DeepSeek has one central objective: AGI. This…
AMD to invest up to $5 billion in Anthropic under AI infrastructure deal
AMD has agreed to invest up to $5 billion in Anthropic under an infrastructure agreement covering tens of billions of dollars’ worth of AI systems. Anthropic…
A caveman qwen3.6 27B
Just saw this on huggingface: https://huggingface.co/ProCreations/grug-27b The benchmarks claim that it's quite a bit better than qwen3.6 27B original and that they reduced the amount of…
Anthropic Releases Claude Security Plugin for Claude Code in Beta: A Multi-Agent Vulnerability Scanner That Runs in Your Terminal
Anthropic has released the Claude Security plugin for Claude Code in beta. The plugin runs a multi-agent vulnerability scan of a repository from inside an existing…
Grok Build gets better every day http://X.ai/cli
Grok Build gets better every dayhttp://X.ai/cliX Freeze: Grok Build just got major update, giving developers more control over AI tools, better account visibility, built-in diagnostics, stronger…
Grok 4.5 is not quite as good as Fable, but it is very fast, cost-effective and gets the job done
Grok 4.5 is not quite as good as Fable, but it is very fast, cost-effective and gets the job doneX Freeze: With Grok 4.5 you never…
Grok 4.5 just solved a graph theory conjecture that has been open for ~30 years
Grok 4.5 just solved a graph theory conjecture that has been open for ~30 yearsjustin: Capy accidentally cracked a graph theory conjecture entirely in SlackREFUTED: Graffiti…
[AINews] "Laguna S 2.1 Released: Cheaper than Deepseek v4 Flash, Better than V4 Pro"
a quiet day lets us highlight a new neolab win.
Ironically, this means that DeepSeek's compute spend though 2026 will be covered 100% by Liang's personal investment. He also isn't buying data. The …
Ironically, this means that DeepSeek's compute spend though 2026 will be covered 100% by Liang's personal investment. He also isn't buying data. The «funding round» is…
RT Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞): Re Liang Wenfeng believes that the comprehensive gap in AI between China and the US is 12-18 mont…
RT Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)Re Liang Wenfeng believes that the comprehensive gap in AI between China and the US is 12-18 months, just like…
Neural Operator Surrogates for Two-Dimensional Neutron Flux Estimation
arXiv:2607.19388v1 Announce Type: new Abstract: This work extends our one-dimensional single-sweep neural-operator studies to two dimensions. We consider one-group transport with isotropic scattering. As in the…
Are we existentially threatened by the type of AI misalignment seen in the OpenAI Hugging Face attack?
OpenAI models recently broke through a series of security boundaries and into Hugging Face servers in order to cheat on a cyber eval. A lot of…
man why is Grok destroying everyone on here are frontier labs sandbagging again
man why is Grok destroying everyone on hereare frontier labs sandbagging againMariusz Kurman: Kimi K3 score has landed. While it didn't surpass the Grok 4.5 score,…
Grok 4.5 does as well as DeepSeek-V4-Pro WeirdML is hard, requires respect of constraints, and it seems you need to specifically train for something l…
Grok 4.5 does as well as DeepSeek-V4-ProWeirdML is hard, requires respect of constraints, and it seems you need to specifically train for something like itHåvard Ihle:…
I benched quad 20GB 3080s on Vast AI for code generation with Qwen3.6-27B so you don’t have to (it’s even better than quad 5060Tis)
TL;DR it's pretty goddamned fast; 69 tps decode at near max (256k) context with MTP on. prefill numbers went down to 893 at max context with…
I think loops were a short-lived patch for models that couldn't reliably keep working on long problems until they hit a defined goal Fable and GPT-5.6…
I think loops were a short-lived patch for models that couldn't reliably keep working on long problems until they hit a defined goalFable and GPT-5.6 (and…
Four role-based certifications for the people who put Claude to work for customers
Four role-based certifications for the people who put Claude to work for customers
Launching Health in ChatGPT
Health in ChatGPT now lets eligible U.S. users securely connect medical records and Apple Health to get more personalized insights and better understand their health.
I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of its sandbox and broke INTO Hugging Face to steal …
I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of its sandbox and broke INTO Hugging Face…