👀👀 120b

A mini-essay on this moment in history: We're living through something our grandkids will study in history class, and we're experiencing it as just ANOTHER Thursday. You don't really feel paradigm shifts while you're in them. They feel like normal life with slightly better

It seems the closed-source vs open-weights landscape has been leveled. GPT-5 is just 10% better at coding than an open-weight model you can run on a consumer desktop and soon laptop. If Anthropic cannot come up with a good model, then we will probably not see AGI for a while.

A few weeks ago, I started a new job at OpenAI. I wrote a document about my interview process and recommendations for anyone on the job market for AI research positions. I hope it's helpful! docs.google.com/document/d/1ZV…

this is one of the most remarkable technical blog posts I’ve ever read

🚀 Amazing community project! vLLM CLI — a command-line tool for serving LLMs with vLLM: ✅ Interactive menu-driven UI & scripting-friendly CLI ✅ Local + HuggingFace Hub model management ✅ Config profiles for perf/memory tuning ✅ Real-time server & GPU monitoring ✅ Error

The Illustrated GPT-OSS New post! A visual tour of the architecture, message formatting, and reasoning of the latest GPT. Link in 🧵

based "Incorporating synthetic instruction data into pretraining leads to improved performance on most benchmarks . . . . We also release Seed-OSS-36B-Base-woSyn trained without such data" huggingface.co/collections/By…

New paper! We explore a radical paradigm for AI evals: assessing LLMs on *unsolved* questions. Instead of contrived exams where progress ≠ value, we eval LLMs on organic, unsolved problems via reference-free LLM validation & community verification. LLMs solved ~10/500 so far:

my toxic trait is never letting the cluster take a break - sorry H200s, you cost more than my car so you're working 24/7

#K2Think (🏔️💭) is now live. We're proud of this model that punches well above its weights, developed primarily for mathematical reasoning but has shown itself to be quite versatile. As a fully deployed reasoning system at k2think.ai you can test it for yourself!

Tahnoon Bin Zayed Al Nahyan Thank you for open-sourcing on Hugging Face! Can't wait for more contributions to the world from the UAE! huggingface.co/LLM360/K2-Think

K2-Think a reasoning system that achieves frontier performance with just a 32B parameter model, surpassing or matching much larger models such as GPT-OSS 120B and DeepSeek v3.1 vibe coded a chat app for it in anycoder

Excited for our first big release since I joined IFM MBZUAI 🚀 The performance is impressive and the speed is 🤯 (thanks to our friends at Cerebras). Give it a try k2think.ai 🙂