This experiment shows how easy it is to poison an open-weight AI model for under $100

A cybersecurity researcher poisoned an open weight AI model for under $100 in about an hour, exposing how easily these increasingly popular systems can be secretly compromised without detection.

Brex built its AI agent policy by watching what agents actually do, not by writing rules first

Brex's CrabTrap proxy inspects AI agent network traffic and drafts security policy from observed behavior, using an LLM judge ...
Tech Times

AI Agents Ported Tao’s 27-Year-Old Math Code in Hours and Found Two Bugs He Had Missed

AI coding agents ported Fields Medalist Terence Tao’s Java 1.0 math visualizations to JavaScript in hours, identified two ...
The Gazette

Russia has looted thousands of Ukrainian cultural objects in the war. Finding them is a challenge

KYIV, Ukraine (AP) — When Alina Dotsenko returned to her museum after Ukrainian forces retook the southern city of Kherson from Russian forces in late 2022, she found thousands of artworks had ...
C&EN

Do-It-Yourself 5-Color 3D Printing of Molecular Orbitals and Electron Density Surfaces

Article Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to ...
C&EN

From Abstract to Manipulatable: The Hybridization Explorer, A Digital Interactive for Studying Orbitals

As digital educational media use becomes more widespread, an opportunity exists to develop new methods to present abstract ideas to provide a more meaningful learning experience. Drawing from ...