-
Stable-GFlowNet uncovers more hidden weaknesses in generative AI modelsKAIST researchers have developed a safety verification technology that uncovers roughly seven times more hidden vulnerabilities in AI than existing methods. The technology is expected to serve as a foundation for developing safer, more trustworthy AI. The work is published on the arXiv preprint server.... Read more -
Standardized battery database could improve AI-guided solid-state battery designThousands of peer-reviewed papers on battery devices are published each year. Despite this rapid growth, the data contained within these publications are often reported in different formats and with varying levels of detail and accessibility. The resulting patchwork of information makes it difficult for scientists, engineers, industry partners and policymakers... Read more -
Claude AI finds cryptography weaknesses human experts missedAI company Anthropic has announced that its unreleased model, Claude Mythos Preview, has discovered previously unknown mathematical weaknesses in cryptographic algorithms that human researchers had missed for years.... Read more -
Strike a pose: Creating more realistic multi-person imagesGenerating an image of a person on a computer using text prompts is easy. Generating one with two people is similarly simple. But creating an image of multiple people actually doing something, and faithfully reproducing not just the people but the thing they're doing? Not so easy.... Read more -
OpenAI says rogue AI agent attack hit other companiesChatGPT maker OpenAI has revealed that an autonomous artificial intelligence agent that hacked a popular platform for computer programmers also attempted to breach four other companies during the incident.... Read more -
Game-engine forests train drone AI to count trees with far less labelingA drone swoops low over an alpine forest. It climbs suddenly to follow the contours of the sharply rising landscape. Pulses from its lidar—a laser mapping instrument—rapidly scan the trees below.... Read more -
What really happens to your data when you click 'delete'?Deletion is a common part of modern life. We send files and folders to the recycle bin all the time and often get rid of unwanted personal accounts. But do you know what really happens when you hit the "delete" button?... Read more -
AI industry slowdown may be needed after security scare, leaders warnOver 1,000 employees at cutting-edge AI companies, including Anthropic CEO Dario Amodei, signed a petition Tuesday calling on the U.S. government to help slow the release of the most advanced AI models.... Read more -
Brain-inspired AI is capable of flexible planning and problem-solving while using far less energyThe capabilities of large AI systems are constantly improving, but they consume a great deal of energy during training and operation. The human brain, by contrast, is extremely energy-efficient: It requires only around 20 watts.... Read more -
Reconfigurable AI device preserves long-term memory while acoustic waves reset short-term signalsA research team led by professor Il Jeon of the Department of Nano Engineering and the Sungkyunkwan Advanced Institute of Nanotechnology (SAINT) at Sungkyunkwan University, including Dr. Sihyeok Kim and Dr. Jang Woo Lee, has developed a next-generation artificial intelligence semiconductor device capable of independently implementing long-term and short-term memory... Read more -
Computer scientists design a cyberattack to prevent cyberattacksProtecting sensitive, precision-timed computing systems requires understanding the nature of cyberthreats. So researchers from Washington State University teamed up with experts at the University of Colorado Colorado Springs and Metro State University to design an attack—dubbed NosyNeighbor—that lurks on the sidelines of time- and safety-critical computer systems, "infers" what's happening... Read more -
Study highlights federated and reinforcement learning for natural language processingNatural language processing (NLP), a branch of artificial intelligence, enables computers to understand and generate human language. Large language models (LLMs) now power everyday applications such as chatbots, virtual assistants and translation tools. Despite these advances, modern NLP systems still face important challenges, including protecting user privacy, adapting to changing... Read more -
Made by AI? EU tells firms to stick a label on it from SundayDeepfakes and other AI-generated content must be labeled from Sunday, as the European Union's sweeping artificial intelligence transparency rules kick in.... Read more -
Workplaces look for cheaper AI as 'tokenmaxxing' fades as a corporate fadA corporate fad of "tokenmaxxing" on artificial intelligence technology is hitting its limits as workplaces throwing AI at everything are seeing the costs rise without a similar spike in productivity.... Read more -
AI gains a tool to identify Kinyarwanda propaganda, with promise for 600 Bantu languagesA new study led by Fabrice Niyigaba '27 reports the first digital tool for identifying online propaganda in Kinyarwanda, the national language of Rwanda's 15 million people—and possibly the first for any of the Bantu languages spoken by 350 million Africans.... Read more
