#LLM
Curated coverage on «LLM»: releases, research, benchmarks, open-source.
SGLang spins out as RadixArk at a $400 million valuation
SGLang, which grew out of a UC Berkeley lab, is now RadixArk, an independent company valued at $400 million and backed by Accel.
Fund Portfolio Shifts: A Focus on AI and Safety
Major Chinese funds are reshaping their portfolios, with a focus on AI, nonferrous metals and stability. Details in the article.
Philippines to lift Grok ban after xAI commits to fixes
Philippine authorities are ready to lift the block on xAI's Grok chatbot after the company promised to take steps to address the violations.
OpenAI launches Edu for Countries to help governments modernize education
Higgsfield: GPT-5 and Sora 2 Turn Ideas into Cinema
WaveFormer: A New Approach to Computer Vision from Peking University and Tsinghua University
R1 anniversary: DeepSeek's Model 1 quietly appears on the horizon
Medical SAM3: Text-Prompted Segmentation Comes to Medical Imaging
Medical SAM3 segments medical images from text prompts alone, which could simplify and speed up diagnosis.
AI for Science breakthrough: inverse design runs 128 times faster
Chinese researchers have sped up the inverse design of multiscale structures with AI for Science, reporting a 128-fold gain in algorithm speed.
Claude Code: A Mac App in 8 Hours – Magic or Grind?
A developer shares what it was like to build a Mac app with Claude Code. Rapid development or painstaking work?
Claude Cowork: A Revolution from Anthropic, or a Pretty but Empty Interface?
A look at Claude Cowork: is the hype justified? We break down what Anthropic's new tool does and how useful it really is. Spoiler: it's complicated.