All news
10783
NEXUS: Runtime Safety Control for Tool-Using LLM Agents — F1 0.949
Researchers introduce NEXUS, a monitor that decides in real time whether to allow, block, or query an AI agent's action, adding less than 0.1% overhead.
Learn2Discern Benchmark: LLMs Barely Tell Reliable Sources from Popular Ones
An arXiv study tested 13 language models on nearly 670,000 trials: they lean on a source's popularity twice as much as on its actual reliability.
FormulaSPIN: A Neural Network That Writes Spreadsheet Formulas From Descriptions and Beats Proprietary Models
Researchers introduced FormulaSPIN, a self-play training framework that generates spreadsheet formulas from text descriptions and outperforms proprietary models…
Apple is preparing an AI shopping assistant: a section about it turned up in the Apple Store privacy policy
Tencent's Marvis: System-Level AI Agent for PCs Hits 300,000 DAU in Two Days
Google Isn't Falling Behind in the AI Race, Pichai Tells Skeptics, Pointing to 82% Cloud Growth
Poolside's Laguna S: a 118B MoE That Beats Thinky's Trillion-Parameter Open-Weights Model
Laguna S 2.1 vs DeepSeek V4: neolab promises cheaper than Flash and stronger than Pro
Lab neolab has unveiled Laguna S 2.1, claiming it is cheaper than DeepSeek V4 Flash and stronger than the flagship V4 Pro. The announcement includes no figures or…
Nunchaku Lite: 4-bit Image Generation in Diffusers — Up to 50% Less VRAM
Hugging Face has integrated Nunchaku Lite 4-bit inference directly into Diffusers: up to 50% less VRAM and image generation about 30% faster, with a single line of…
China Cuts Its AI Gap to 6%: Moonshot and Zhipu Are Not a One-Off
AI Data Center Boom: Investment Applications in Thailand Jump 80% to $40.6 Billion
ByteDance's Seed Edge Invites 100 Scientists to Seed STEM Program for AI-Driven Research
Clobotics' Sparrow Robot Repairs an Offshore Wind Turbine Blade in 27 Minutes
Nvidia gives DGX GB300 supercomputer to US Navy's postgraduate school for free
Nvidia donated a DGX GB300 AI supercomputer to the US Navy's postgraduate school, the first system of its class in the US armed forces. Jensen Huang came to…
Computer Vision in Retail: Lenta Reads Price Tags From a Moving Robot
Lenta turned price-tag recognition from a moving robot into a Computer Vision ML hackathon, aiming to automate shelf monitoring and violation detection in its…