RT @matthewclifford: UK academics: escape GPU poverty! SovAI wants to help… More below👇.

RT @sarahookr: One of the first technical reports in a while that has considerable detail. In particular -- preference training tends to b….

RT @mustafasuleyman: MAI-DxO in action, tackling one of those complex cases:

RT @Dominic1King: Meet MAI-DxO, our AI Diagnostic Orchestrator. By emulating a virtual panel of specialists, it boosts base-model accuracy….

RT @chrisck: New paper today! 🥳 How good is generative AI at diagnosis compared to human doctors? We introduce a novel, interactive medical….

RT @mustafasuleyman: We're taking a big step towards medical superintelligence. AI models have aced multiple choice medical exams – but rea….

It is remarkable that a committee of deliberating AI models can diagnose patients, propose the next tests, use the results of the tests to improve the diagnosis, order more tests, and eventually arrive at diagnoses, outperforming doctors. It is a specific setting, but nonetheless.

RT @StefanoErmon: Huge milestone from the team! . A blazing-fast diffusion LLM built for chat, delivering real-time performance at commerci….

Thanks @MiaoYishu for sharing your thoughts with me on this ❤️.

AI and China: silicon valley AI folks love to talk about the AI race against China. Let’s not talk about China (the abstract entity), but rather the Chinese people. Their achievements have been brilliant and worthy of praise: An exceptional focus on education, hard work and.

RT @chihyaoma: 💼 [Career Update] I feel fortunate to have spent the past 4 years at Meta, working with some of the brightest minds to build….

RT @ArxivToday: This is wild - researchers built a neural network that can control real-world entities like cars, bikes & pedestrians using….

RT @FrankRHutter: It's great to have this new open-source benchmark for tabular data 🚀.It's really comprehensive, created and maintained b….

RT @arankomatsuzaki: jina-embeddings-v4: Universal Embeddings for Multimodal Multilingual Retrieval. - 3.8B embedding model supporting both….

RT @HannesStaerk: Monday Starkly Speaking: we understand a core diffusion model ingredient better - classifier free guidance. Via "Classi….

RT @mathusmassias: New paper on the generalization of Flow Matching 🤯 Why does flow matching generalize? Did you k….

RT @baophamhq: Diffusion models create novel images, but they can also memorize samples from the training set. How do they blend stored fea….

RT @percyliang: Wrapped up Stanford CS336 (Language Models from Scratch), taught with an amazing team @tatsu_hashimoto @marcelroed @neilbba….

RT @lmarena_ai: 🚨Breaking: New DeepSeek-r1 (0528) just tied for #1 in WebDev Arena, matching Claude Opus 4!. More highlights:.💠 #6 Overall….

RT @arankomatsuzaki: Gemini 2.5 tech report was released!