10 results on this page
ByteDance’s Seed foundation-model division has reportedly completed another internal restructuring. Its foundation-model organization now includes four first-level departments focused on pretraining data, reinforcement learning, product post-training for work, and product post-training for chat.
Kuaishou’s Kling AI video-generation business generated more than RMB850 million in second-quarter revenue. Revenue rose over 200% year on year and 30% quarter on quarter, reaching RMB1.5 billion in the first half of the year.
This survey frames self-evolving LLM agents as dynamic graphs whose memories, tools, skills, workflows, and relationships change over time. It presents four evolution taxonomies, connects nine dynamic-graph-learning fields to agent capabilities, and proposes graph-aware evaluation and governance protocols.
A study found that LLM recommendations could persuade evaluators to reject promising innovations or approve weak ones. Providing narrative explanations made participants more likely to follow incorrect AI decisions, while unexplained recommendations led to better independent judgment.
Apple Machine Learning studies how to mix scarce target-domain data with abundant generic data during language-model pretraining. Across more than 2,000 training runs, the research examines the trade-off between insufficient target exposure, repeated examples, diminishing returns, and overfitting.
NVIDIA demonstrates how developers can build real-time edge AI applications with Holoscan using its CLI, skills, documentation, and AI coding agents. The workflow leverages HoloHub’s reference applications and reusable components to help agents assist with development.
An MIT study finds that AI-generated images often cannot be traced to individual training images. As datasets grow larger, the influence of any specific example tends to fade.
The headline claims that OpenAI suddenly halted GPT-6 training because of safety concerns. The supplied excerpt offers no confirmation, technical evidence, or explanation of the alleged pause, so the claim should be treated as unverified.
NVIDIA FLARE enables organizations to coordinate training for vision-language models while keeping sensitive image and text data at local institutions. The approach supports multimodal use cases such as visual question answering, captioning, and image-text reasoning without centralizing raw records.
Silicon Data is a startup building tools to help Wall Street assign prices to AI compute. Its goal is to make compute costs easier to value and give firms ways to hedge against price changes as data center and GPU spending expands.