AI
Disrupting malicious uses of AI: June 2025
OpenAI has released a report detailing its efforts to detect and prevent malicious uses of artificial intelligence. The report includes case studies o...
AI
AI
KV Cache from scratch in nanoVLM
Researchers have implemented a key-value cache in the nanoVLM model, which is a lightweight version of the popular VLM (Vision-Language Model) archite...
AI
Real-Time AI Sound Generation on Arm: A Personal Tool for Creative Freedom
Researchers have developed a real-time AI sound generation tool that can run on ARM-based devices. This allows users to generate high-quality sounds i...
AI
Holo1: New family of GUI automation VLMs powering GUI agent Surfer-H
Holo1 is a new family of visual language models (VLMs) designed for GUI automation. These VLMs are intended to power the Surfer-H GUI agent, which can...
AI
SmolVLA: Efficient Vision-Language-Action Model trained on Lerobot Community Data
Researchers have developed a new AI model called SmolVLA, which can process both visual and language inputs to perform actions. The model is trained o...
AI
No GPU left behind: Unlocking Efficiency with Co-located vLLM in TRL
Researchers have proposed a method to improve the efficiency of large language models by co-locating them on multiple GPUs within a single node. This ...
AI
Operation “Sneer Review”: China-origin influence activity
OpenAI has taken action against malicious use of its technology. The company banned accounts that used artificial intelligence to generate negative po...
AI
Operation “Wrong Number”: AI-assisted task scam
OpenAI has taken action against a group of users who were using its platform to facilitate scams. The users, likely based in Cambodia, employed AI too...
AI
STORM-2035: Recidivist influence activity
OpenAI has taken action against a suspected Iranian operation called STORM-2035. The group allegedly used artificial intelligence to generate content ...
AI
Operation “Uncle Spam”: US polarization influence activity
OpenAI has taken action against a group of Chinese-origin accounts that were using AI to create and spread US-focused propaganda. These accounts gener...
AI
Operation “High Five”: Philippines political comments
OpenAI has taken action against accounts generating malicious comments on Philippine politics and public figures. The company's 'High Five' operation ...
AI
Operation “Helgoland Bite”: German-language influence activity
OpenAI has taken down accounts it believes were created by Russian actors. These accounts used AI-generated content to spread misinformation on Ukrain...