Vision Transformers

🚀 We’re excited to introduce Qwen3-235B-A22B-Thinking-2507 — our most advanced reasoning model yet! Over the past 3 months, we’ve significantly scaled and enhanced the thinking capability of Qwen3, achieving: ✅ Improved performance in logical reasoning, math, science & coding…

New video on the details of diffusion models: youtu.be/iv-5mZ_9CPY Produced by @welchlabs, this is the first in a small series of 3b1b this summer. I enjoyed providing editorial feedback throughout the last several months, and couldn't be happier with the result.

the level of incompetence in this post is so confusing, i mean why would you use a transformer here in the first place?

230k GPUs, including 30k GB200s, are operational for training Grok @xAI in a single supercluster called Colossus 1 (inference is done by our cloud providers). At Colossus 2, the first batch of 550k GB200s & GB300s, also for training, start going online in a few weeks. As Jensen…

India has around 5,000 H100 GPUs in total. Elon Musk alone has 150,000. Let that sink in. We're not just behind in the AI race; we're not even on the track. At this rate, only a miracle can pull us back into the game. And if we lose this race now, forget about catching up…

Agents aren’t reliable. They don’t learn from experience. At @composiohq, we provide skills that evolve with your agents @lightspeedvp gave us $25M to make agents usable

AI researchers when they discovered that torch.compile doesn't scale well to real multi-node production training workloads and is a giant footgun

Excited to share Aryabhatta 1.0, our leading model that scores 90.2% on JEE Mains, outperforming frontier models like o4 mini and Gemini Flash 2.5 Trained by us at @AthenaAgentRL , in collaboration with @physics__wallah, using custom RLVR training on 130K+ curated JEE problems…

🚨 Olympiad math + AI: We ran Google’s Gemini 2.5 Pro on the fresh IMO 2025 problems. With careful prompting and pipeline design, it solved 5 out of 6 — remarkable for tasks demanding deep insight and creativity. The model could win gold! 🥇 #AI #Math #LLMs #IMO2025

This past week, Harmonic had the opportunity to represent our advanced mathematical reasoning model, Aristotle, at the International Mathematics Olympiad - the most prestigious mathematics competition in the world. To uphold the sanctity of the student competition, the IMO Board…

got rescinded from columbia

took a quick look at this paper (just the convolution section) and I have several concerns about the claims: 1) pytorch by default does not execute synchronously on the GPU (host vs. device) and anyone who has forgotten syncs when benchmarking can tell you so 2) TF32 is enabled…

CUDA-L1: Improving CUDA Optimization via Contrastive Reinforcement Learning Trains a DeepSeek-v3-671B model to optimize CUDA kernels using only execution-time speedup as reward. Pipeline: - SFT: Finetuned on 2.1K correct, executable CUDA variants from 6 LLMs across 250…

In this episode of AI Talks by AIM powered by @Meta with @sunil_abraham , Public Policy Director - Data Economy and Emerging Tech at Meta, India, we dive into a powerful conversation about how open-source generative AI is enabling real-world impact, far beyond Silicon Valley.…