MarkTechPost (57 minutes ago)
0sec Labs has open-sourced 0, an AI security harness for code review, pentesting, reverse engineering and fixes. It runs locally through a browser console, a headless CLI and an MCP server, with your own model provider.
Read the full story on MarkTechPost.
Working with cryptographers, DeepMind’s AI agents cut the cost of forging RSA-1024 signatures and found a cheaper way to factor keys.
Rillet, a US start-up that makes accounting software with built-in AI agents, has opened an office in central London. It is the company’s first major move into Europe, Rillet said on Monday.
Pinterest has moved its user-journey inference system from a multi-stage clustering pipeline to a fine-tuned 4-billion-parameter language model, reporting a 1.1% improvement in email click-through and 1.3% higher push opens. By Matt Foster
The OpenAI engineer who was arrested and charged last month for planting a covert audio recording device in the company’s office has pleaded not guilty.
Tapping into the country’s strengths in manufacturing and engineering, NVIDIA is helping advance Korea’s physical AI ecosystem. The country’s industry leaders are applying accelerated computing, AI and simulation to everything from semiconductor manufacturing to robotics and autonomous vehicles.
As AI agents take on more work, they need capable open models and security that advances alongside them.
Google’s Elephant-Goldfish Model plugin makes coding agents draft specs, then runs 5 zero-context review stages before any code.
A config-verified comparison of how 10 open MoE models route tokens: total and active parameters, expert counts, top-k, shared experts, gating functions and load-balancing methods.
MoE routing strategies decide which experts process each token. This guide explains 7 methods, from Mixtral’s top-2 softmax to DeepSeek-V4’s hash routing, with the models that use each one.
Nous Research released Arete-27B, an Apache 2.0 fine-tune of Qwen3.8-27B. On Nous’s 803-prompt eval, propaganda answers fell from 22% to 0.9% while Hermes Bench stayed flat at 65.60%.
69 active sources
24 Turkish · 45 English
All sources
