Microsoft’s multi-agent AI system tops Anthropic’s Mythos on cybersecurity benchmark

Microsoft's new vulnerability-scanning system, codenamed MDASH, scored 88.45% on the CyberGym benchmark, surpassing single-model systems from Anthropic and OpenAI by using more than 100 specialized AI ...
SpaceNews

Benchmark Space Systems Fires Up Metal Plasma and Bi-Prop Thruster Production, Realigns Senior Executive Team to Deliver on Intensifying Hybrid Propulsion System Demand

Laser focused on delivering mobility without compromise, Benchmark Space Systems today announced it has signed contracts for nearly two dozen new electric metal plasma thrusters (MPTs), with some set ...
SpaceNews

Benchmark wins $4.9 million award for ASCENT propulsion systems

SAN FRANCISCO – The Air Force Research Laboratory awarded Benchmark Space Systems $4.9 million to develop propulsion systems for ASCENT monopropellant. The two-year award announced Sept. 5 covers ...
Business Wire

New MLPerf Inference v4.1 Benchmark Results Highlight Rapid Hardware and Software Innovations in Generative AI Systems

SAN FRANCISCO--(BUSINESS WIRE)--Today, MLCommons® announced new results for its industry-standard MLPerf® Inference v4.1 benchmark suite, which delivers machine learning (ML) system performance ...
MIT Technology Review

How to build a better AI benchmark

To fix the way we test and measure models, AI is learning tricks from social science. It’s not easy being one of Silicon Valley’s favorite benchmarks. SWE-Bench (pronounced “swee bench”) launched in ...