سختافزار
Open From Data Center to Desktop: NVIDIA Expands Nemotron Family With Highest Efficiency Super VL and Bolt Models
NVIDIA today released Nemotron 3.5 Super VL, an open 120-billion-parameter model that delivers the fastest task completion and highest accuracy for specialize
مدلها
OpenAI launches Dots, its bubbly agentic avatar
Dots are meant to operate independent of any specific hardware or interface, pursuing user-defined goals continuously in the background with minimal oversight
محصولات
Humans, not rogue AI, are still the biggest cybersecurity risk to energy systems
Before recent high-profile hacks raised the specter of AI possibly "killing all humans," our energy systems were already disturbingly vulnerable to cyberattac
اخبار
Why Specialization Is Inevitable
A Blog post by Dharma-AI on Hugging Face
پژوهش
هوش مصنوعی در تولید متن سریعتر: Self‑Speculative Decoding
بهبود سریعسازی تولید متن با استفاده از Self‑Speculative Decoding، که ترکیبی از تحلیل زودهنگام و بازنگری در لایههای LLM است، میتواند سرعت و مصرف حافظه را به طور چشمگیری کاهش دهد.
پژوهش
هوش مصنوعی به سرعت میرود: Dynamic Speculation در Hugging Face
این خبر، فناوری Dynamic Speculation را معرفی میکند؛ روشی اجمالی و سریع برای تولید متن با سرعت تا ۲.۷ برابر، از Intel و Hugging Face. سریع و دقیق.