33 results on this page · clear filters

ai Sep 11 BlueSTAR: Tiered Agentic Architecture for Autonomous Cyber Defense Security analysts are losing the speed war. Average breakout time, from initial access to lateral movement onto a second host, is now 29 minutes, and the fastest observed is 27 seconds. Meanwhile, defenders still rely on human jud... ai Sep 10 Anthropic CEOs wife once asked Epstein to fund porn venture – now steers Claude The woman who helps steer Anthropic, the AI company racing toward a potential $2 trillion valuation, spent two years cultivating a relationship with Jeffrey Epstein while pitching him on a "luxury porn" venture. The connection, re... ai Sep 10 Mathematicians want proof OpenAI didn't use their work Andreas Thom, a mathematician whose work on non-sofic groups became one of OpenAI's ten announced breakthroughs, is demanding the company prove it did not use his unpublished research to train its models. His accusation, posted on... ai Sep 10 Training a 3.8B LLM to 0.384 CORE for $998 – Hugo Vergnes Most language model training happens in two zones: tiny experiments on a single GPU, or massive runs that cost millions. Hugo Vergnes wanted to explore the space between them. His result is a 3.8B-parameter model trained from scra... ai Sep 10 JarvisGUI: Towards Cross-Device GUI Agents with Dynamic Task Composition When you copy a file from your phone to your laptop, open it in a spreadsheet app, tweak a formula, email the result to a colleague, and ask them to check it on their desktop, you are performing a cross-device GUI workflow. It inv... ai Sep 07 The smallest edge AI device for local LLMs Tiiny AI wants to put a 120-billion-parameter language model in your pocket. The company, founded in 2024 by a team from MIT, Stanford, HKUST, SJTU, Intel, and Meta, debuted the Tiiny AI Pocket Lab at CES 2026 and secured a Guinne... ai Sep 07 Tell HN: OpenAI brings back 5 hour limit for plus and business standard users OpenAI appears to have reinstated usage limits that restrict Plus and Business Standard users to a five-hour window, a constraint that had been lifted previously. The change was flagged by a Hacker News user who noticed the limits... ai Sep 07 Speculative Decoding in vLLM on AMD GPUs vLLM has added support for speculative decoding on AMD GPUs, a technique that lets the serving system verify multiple drafted tokens in a single pass through the target model. The feature works across five different drafting metho... ai Sep 07 Nvidia's Jensen Huang says 'AGI has arrived' and congratulates OpenAI Nvidia CEO Jensen Huang declared on Sunday that artificial general intelligence has arrived, congratulating OpenAI on the release of Astra, the company's newest and most powerful model. The statement, posted on X, adds another voi... ai Sep 07 Same Trajectory, Contradictory Rewards (ROBORMBENCH): Paraphrase Fragility in Vision Language Reward Models Vision-language models are being adopted as reward functions for robotic learning at an accelerating pace. The idea is straightforward: feed a robot trajectory and a text description of the goal into a VLM, and the model outputs a...