Knowledge catalogue
Translation: 'Internal beta testing for an intermediate version of DeepSeek V4.1 Flash is now open; you are welcome to try it out. It adopts a new model architecture featuring native multimodal suppor
Siobhan Roberts / New York Times: Fields medalist Jacob Tsimerman, set to join OpenAI later in September, launches the Mathematical AI Safety Institute to apply higher math to AI safety problems — Jac
I've been making a lot of comments about optimal setup for Strix Halo (gfx1150) and from my observation, 90% of our community is using offcial llama.cpp for it, which is NOT optimized for Strix Halo a
@GaryMarcus was right. It was the harness, not the model. That’s a much bigger deal (and hidden danger) than the benchmark score. If the real capability lives in opaque, vendor-controlled scaffolding
A few weeks ago, I saw this post where the author trained a SIREN MLP to implicitly memorize Bad Apple as a coordinate function: (t, y, x) to pixel. That got me curious about a slightly different form
Google LLC’s DeepMind research unit said today it has used its AlphaGenome artificial intelligence model to predict the biological consequences of the more than 9 billion possible single-letter change
Threat actors used a multi-agent artificial intelligence framework to compromise thousands of credentials in under six hours, Google LLC’s Google Threat Intelligence Group said in a report released to
Executive Summary Since the release of our May 2026 report detailing adversarial misuse of artificial intelligence (AI), Google Threat Intelligence Group (GTIG) has observed forward leaning adversarie
OpenAI announced a quick-start solution for creators who have an idea but need help turning it into visual content: use ready-made templates for popular formats such as posters or merchandise and then
When building consumer-facing generative AI applications, balancing high generation quality with fast response times across diverse media types, can be challenging. KDDI, a major telecommunications ca
I'm the CTO of a mid-size nonprofit, and I run a local Qwen stack as a coding subagent for that work — partly on principle (a fair amount of what I handle shouldn't leave the building), partly because
I am pleased to see that OpenAI’s new model is roughly on par with Gemini Flash and Opus 4.8 on prompt injection risk. Nice work! Evaluating and naming other labs turns out to be a great way to encour
I kept picking local models badly. Parameter count, quantization, VRAM. I could read all of it and still not know whether a model would be any good on my machine. So I stopped guessing and wrote somet
Hello LocalLLaMA Fam! I built this project because I, like many of us here, enjoy local AI and have a nice setup, but still find myself reaching for Cloud-based AI due to their convenience. My models
I found a lot of room on the table for these cards so I decided to make a specialized build to squeeze all I could. first The results: qwen 3.8 next Q3_K_XL: 920tk/s pp8192 (2 cards, ram offload), 24/
i once asked a bunch of deep-tech researchers on why they wouldn’t use chatgpt for serious research, especially after openai launched Prism tool around a workflow researchers already know well: LaTeX.
I'm testing a new approach for reducing the cost of image-based LLM inference. I evaluated it on the MOMA Graph benchmark, using 1,315 questions. Compared with using GPT-4o to process the original ima
Hayden Field / The Verge: In an expanded class action lawsuit, a group of Claude subscribers says Anthropic deceptively advertised the limits of its Max subscription tier — Anthropic says power users
Ling-3.0-flash-VL inherits the language, reasoning, and long-context capabilities of Ling-3.0-flash, while extending them with native image and video understanding. The model has 124B total parameters
Introducing ChatGPT Images 2.5 OpenAI's image generation models are apparently used 'more than 3 billion images across ChatGPT Images and the GPT‑Image models in the API'. This latest release improves
I’m currently running llama.cpp on two RTX 2060 12GB cards, so 24GB total VRAM. With Qwen3.8 27B IQ4_XS at 131k context I’m getting around 45 tok/s, which is actually pretty good for this setup. I fou
Hey so a few months ago I got into local LLMs and didn’t really like the tools out there at the time, so I decided to build my own version of a Claude/Codex GUI for local models. I’ve been working on
Ina Fried / Axios: Meta launches Muse, a personal AI agent that runs on a dedicated VM in Meta's cloud, initially available in the US, with support coming soon for its AI glasses — Meta on Tuesday ann
Adam Satariano / New York Times: Mistral raised a €3B Series D led by Samsung at a €21B valuation, up from €11.7B a year ago, as it expands into data centers beyond developing AI models — Mistral is t
So I've been messinga round with embedding models for a bit, and I think they are interesting enough to experiment with. They are useful for rag, especially in a localllm sense because you can ground
Nice report on agents beyond code generation. Here is why it matters: Coding agents raise how much code gets written. This report argues the gains shrink sharply between writing code and shipping reli
Nvidia CEO @JensenHuang says GPT-6 Astra means “AGI has arrived.” @GaryMarcus says that is false. He points to the same AGI declarations around GPT-5 and o3, neither of which delivered what was promis
Hi everyone, This is probably a simple problem, but I can't get my head around it. I am running ollama 0.33.3 on Linux Mint Cinnamon 21.3 with separate partitions for root and home. The standard insta
On the Navier–Stokes Millennium Prize Problem Impressive result from OpenAI, who used an unreleased model to produce a resolution to the Navier–Stokes existence and smoothness problem, one of the seve
I was reading up on the recent controversy around Tristan Buckmaster, Levent Alpöge, OpenAI, and the Navier–Stokes result, and it got me thinking about something broader than this particular dispute.
French artificial intelligence lab Mistral AI SAS today announced that it has raised €3 billion, or about $3.49 billion, in funding. Samsung Electronics Co. Ltd. led the investment. It was joined by m
Ina Fried / Axios: OpenAI launches ChatGPT Images 2.5, which it says cuts image generation latency by up to 50% vs. Images 2.0, and adds a Sketch feature for drawing in ChatGPT — OpenAI is launching C
Other investors in the round include Benchmark, Bessemer, Kleiner Perkins, Greylock, Lightspeed, Altimeter, Bond Capital, Meritech, Atreides, Valor, T. Rowe Price, Lux, 8VC, D1, 137 Ventures, DST, Pos
Our open-weight models, products and infrastructure give organisations a real choice over how and where they run AI, not just access to a model. That's frontier performance without the lock-in. Full s
Our partners at @CosineAI train their own coding model on real production code, for languages like Fortran and Verilog that generic models fumble. On cost per successful task, their Lumen Outpost mode
Pathway's Baby Dragon Hatchling (BDH) is a brain-inspired, post-transformer architecture that reasons in latent space instead of emitting chain-of-thought tokens. See how Pathway develops and scales B
Enterprise agent adoption isn’t one-size-fits-all. While many teams will opt for managed commercial platforms, such as Gemini Enterprise Agent Platform for turnkey agent deployment and governance, dev
practically my entire feed for days had consisted of string of reports about OpenAI’s misconduct, around the hacks, around apparent coverups, possible data fudging and maybe even intellectual theft wi
I was inspired by Bijan Bowen video - Subway FPS https://youtu.be/6kjXzTVmT58?t=1035 Wondered how far I can push Qwen 3.8 27b so I used a plan made by Fable 5.1 DESIGN.md which has 267 KB! ( 26K of de
Up front: I'm one of the people building the page-perception layer used here. We started by testing small local models. The result turned out to be more interesting than the original test. 12 small mo
Hey guys, After my CPU-only to 96GB VRAM test, I tested Qwen3.8-Flash-Next across llama.cpp, SGLang and FreeToken on the same workstation. This time I wanted to see what changes when you keep the hard
An update to my previous post on running Flash-Next with the expert cache and MTP. I found another useful improvement on the same dual-3090 setup: replacing the CUDA top-k fallback used by my build. O
I don't think anyone posted about this here, but Qwen released a finetuned version of 3.5 4 for driving. The full Bf16 checkpoint is 9B. This is a very interesting development of Chinese AI labs tackl
so it was the harness not the model. more sneaky games with unacknowledged domain-specific neurosymbolic AI doing the real work. OpenAI published 99.9%. The benchmark's own software returned 62.7%. Sa
The group produced an analytical proof and Lean formalization that via Navier-Stokes dynamics a fluid can develop a singularity in finite time. The solution is a vortex, a spinning swirl of fluid, tha
Reuters: The NSA, CISA, and FBI issue a joint advisory warning that Chinese AI companies, including DeepSeek, are conducting “industrial-scale” distillation campaigns — The U.S. government on Tuesday
This is legitimately terrible for OpenAI --- but also Anthropic --- and it relates to their work in drug discovery. I read this as: prompts will be monitored and potentially acted on if we think there
Mistral AI announced on 8 September 2026 that it had closed a €3 billion Series D financing round. The investment represents the largest equity round ever raised by a European technology firm, and com
Ok I know I'm likely to get steamrolled here but I also have a sense of optimism and hope I can get some sort of answer that will be beneficial. We all know since the new price structure, usage has sh
Gemma 4 12B runs on an RTX PRO 4500 Blackwell. Gemma 4 E2B run on a Jetson Orin NX 16GB; similar performance is expected on a Jetson Orin Nano Super 8GB. Both systems use a reSpeaker Flex 4-mic array
We congratulate Levent Alpöge and Tristan Buckmaster on their remarkable mathematical work. We (the researchers and the agents) did not see any of their work through any means until they released it p
We have just raised 3B€ to accelerate our R&D. We are scaling up our training and inference compute to build future generations of our models, and expanding our Science team across all our offices --
We want to celebrate with people using GPT-6. We did this for GPT-5.5 and it was really fun. We’re getting together in SF on September 16 to talk about the model, what we should build next, and mostly
Weakly General AI achieved (at least according to the criteria set from 2020). a year before recent predictions ✅Loebner prize was a weak Turing Test, equivalent achieved by GPT-4.5 ✅Winograd passed b
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-ge
We've never seen this before. The biggest jump in Vending-Bench history. GPT-6 Astra is better at making money and more ethical than Claude Fable 5.1. Surprising, because: 1. First time ever that Open
I'm looking for more information to expand my creativity around this. I don't really have a realistic idea of what people actually do with local AI yet, I mostly just want to explore the possibilities
Setup, so the question makes sense: my coding agent's memory now lives in a folder of Markdown — entities, episodes, and procedures (workflows with a per-step success/fail record). Nothing needs an ac
Quick setup on linux: STEP 0: install/download llama.cpp, Freecad, your favourite gguf model - possibly with multimedia image reading capabilities (I've used Qwen3.8-27B-UD-Q4_K_M and relative mmproj-
arXiv:2609.04539v1 Announce Type: new Abstract: A critical challenge in deploying Large Language Models (LLMs) is developing reliable mechanisms to estimate their confidence, enabling systems to deter