Multimodal AI - Page 16
Stories from Multimodal AI come first, newest first, then stories from connected topics (open weights, AI Models, Hugging Face, AI, Nvidia, AI Agents), each labelled with where it comes from. This view is not indexed by search engines.
Latest
MLCommons Ranks 27 AI Systems on Safety; Open Models Span 4th to 98th
MLCommons' AILuminate Safety 1.2.1 puts 22 open-weight models between the 4th and 98th percentile. Jailbreak 1.1 scores follow after developers are told.
OpenAI Rebuilds Its Model Spec Around Five Core Values for GPT
OpenAI has reorganised its Model Spec around five values, adding rules on human oversight and context and dropping most of its style guidance.
USD.AI Leads $106m Loan for NexGen Cloud’s GPUs in Finland
USD.AI, Silver Point and Victory Park lend NexGen Cloud $106m over four years for 1,728 NVIDIA B300 GPUs in Finland, backed by a take-or-pay contract.
Canada to Take Up to C$1.6bn Stake in Cohere as It Raises Series E
Canada will invest up to C$1.6bn in Cohere in stages, as the AI firm opens a Series E targeting up to C$7bn at a C$35.7bn valuation, led by Schwarz Group.
Hugging Face and Five US Universities Launch ProtoStar for Science
Hugging Face and five US universities have launched Project ProtoStar to make publicly funded science data AI-ready, starting with a climate challenge.
Mistral Opens Large 4 Weights: a 1-Trillion-Parameter Model
Mistral has released the weights of Large 4, a 1-trillion-parameter multimodal model, and warns its cyber skills carry a potential for real attack uplift.
Kakao Open-Sources Two Kanana-2 Embedding Models for Korean Search
Kakao has released 3bn and 1.3bn-parameter Kanana-2 embedding models on Hugging Face, topping its Korean retrieval tests while trailing Qwen3 in English.
Particle-1 Coding Model Uses About 45% Fewer Tokens, Particle Says
Suhail Doshi's Particle has launched Particle-1, a DeepSeek V4.1 Flash coding model it says uses about 45% fewer tokens with similar accuracy.
SenseTime Releases Code2Video, an Open Model That Makes Videos in Code
SenseTime's open SenseNova 6.8 Code2Video, a 27B model built on Qwen3.8-27B, writes and renders code to make short videos. It has no benchmarks yet.
PyTorch 2.15 Makes Its Rewritten NCCL Backend the Default
PyTorch 2.15 switches every nccl job to a rewritten backend, drops Python 3.10 and CUDA 12.6 wheels, and adds FP8 on Apple Silicon and NVFP4 on Blackwell.
Latest on TSN
Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.
