AgentLeak introduces the first full-stack benchmark for privacy leakage in multi-agent LLM systems, covering internal channels like inter-agent messages. It spans 1,000 scenarios across healthcare, finance, legal, and corporate domains.
ArXiv AI · 216 天前
Gemini 3 Deep Think receives a major upgrade, achieving state-of-the-art results across domains, especially programming. Only 7 people globally outperform it.
cnBeta (Full RSS) · 216 天前
Gemini 3 Deep Think upgrade achieves SOTA across domains, especially programming where only 7 people worldwide outperform it. This Google VP side project marks a new era in AI reasoning.
cnBeta (Full RSS) · 216 天前
Former OpenAI researcher Zoë Hitzig warns ads in ChatGPT risk user manipulation like Facebook. She left after ad testing amid privacy concerns from user-shared intimate thoughts.
cnBeta (Full RSS) · 216 天前
Former OpenAI researcher Zoë Hitzig quit after testing ChatGPT ads, warning of manipulation risks from users' private data. She compares it to Facebook's pitfalls.
cnBeta (Full RSS) · 216 天前
OpenAI is launching ads on ChatGPT this week amid billions in funding needs. CEO Sam Altman previously opposed ads, calling them a last resort that could erode user trust.
cnBeta (Full RSS) · 216 天前
Google unveiled a major upgrade to Gemini 3 Deep Think, a reasoning model for science, research, and engineering. Google AI Ultra subscribers can access it now in the Gemini App.
cnBeta (Full RSS) · 216 天前
Google announced a major upgrade to Gemini 3 Deep Think, a reasoning model for science, research, and engineering. Google AI Ultra subscribers can access it via the Gemini App.
cnBeta (Full RSS) · 216 天前
AI companies with deep user privacy access are rushing to monetize via ads amid lax regulation. Anthropic's Super Bowl ads satirized OpenAI's vulnerabilities without naming them.
cnBeta (Full RSS) · 216 天前
An AI safety leader warns of global peril and resigns to study poetry. This coincides with an OpenAI researcher quitting over ChatGPT ad testing plans.
BBC Technology · 216 天前
A prominent AI safety leader resigned, warning the world is in peril, to study poetry. This follows an OpenAI researcher's exit over plans to test ChatGPT ads.
BBC Technology · 216 天前
Samsung claims first to ship HBM4 memory, a day after Micron's announcement. HBM4 provides faster, denser RAM for next-gen AI hardware.
The Register - AI/ML · 216 天前
Discusses if agent frameworks remain necessary as LLMs improve. Argues building approaches evolve but agents are fundamentally systems around models.
LangChain Blog · 216 天前
Explores whether agent frameworks are still necessary as LLMs improve. Notes that optimal agent-building approaches evolve with model performance.
LangChain Blog · 216 天前
MiniMax released an updated M2 large language model for real-world productivity. The cheap AI follows rivals' launches in China's intense AI race.
SCMP Technology · 216 天前
MIT report shows frontier models like OpenAI's GPT rely on more computing power rather than smarter algorithms. This scaling approach drives progress but hikes costs.
ZDNet AI · 216 天前
Xiaomi open-sources its first-generation VLA large model for robotics. Part of morning tech news roundup alongside OpenAI updates.
Ifanr (爱范儿) · 216 天前
Security flaws in an AI coding platform allowed a BBC reporter to be hacked. Vibe-coding tools enable non-coders to build apps using AI.
BBC Technology · 216 天前
Cloudflare optimizes websites for AI agents by converting HTML to Markdown. Shifts from blocking bots to attracting them with faster content.
The Register - AI/ML · 216 天前
UJET CEO claims AI enhances call agents into 'superheroes' without job loss. Improves software to resolve issues without multi-system navigation.
The Register - AI/ML · 216 天前
Wesfarmers plans to deploy agentic AI across its entire group. The company has secured multi-year deals with Google Cloud and Microsoft to support this initiative.
iTNews Australia · 216 天前
Robotics companies face major hurdles in developing hands. Creating durable and affordable robotic hands remains a key challenge.
BBC Technology · 216 天前
Enflame Technology started IPO inquiry on Shanghai's Star Market. The AI chip unicorn follows hot semiconductor listings.
SCMP Technology · 216 天前
Apple ML extends μP for hyperparameter transfer across model sizes, modules, width, depth, batch, and duration. Introduces Complete(d) Parameterisation unifying width-depth scaling.
Apple Machine Learning · 216 天前
Apple extends μP for hyperparameter transfer across modules, width, depth, batch, and duration. Introduces Complete(d) Parameterisation unifying width-depth scaling.
Apple Machine Learning · 216 天前
Apple ML advances federated optimization for stochastic variational inequalities (VIs). It provides improved convergence rates closing the gap with federated convex optimization.
Apple Machine Learning · 216 天前
Apple advances federated optimization for stochastic variational inequalities. Establishes improved convergence rates closing gap with convex optimization.
Apple Machine Learning · 216 天前
This paper improves federated optimization for stochastic variational inequalities with tighter convergence rates. It refines analysis for Local Extra SGD on smooth monotone VIs, closing the gap with convex optimization bounds.
Apple Machine Learning · 216 天前
Apple ML paper advances federated optimization for stochastic variational inequalities. It provides improved convergence rates, closing gaps with convex optimization bounds.
Apple Machine Learning · 216 天前
Hugging Face launches custom kernels powered by OpenAI's Codex and Anthropic's Claude. These are now available to all users on the platform.
Hugging Face Blog · 216 天前