GradMAP: Fast Decentralized Learning for Grid-Edge Flexibility
Discover GradMAP, a gradient-based multi-agent learning method enabling efficient, decentralized control of grid-edge energy resources with physics complia...
Layerwise Convergence Fingerprints for LLM Misbehavior Detection
Discover Layerwise Convergence Fingerprinting, a tuning-free method to detect runtime misbehavior in large language models with high accuracy and security.
Google Adds 25M Subs in Q1 via YouTube & Google One
Google gained 25M new subscriptions in Q1, driven by YouTube Premium and Google One, boosting its total to 350M paid users worldwide.
Limits of Automated Evaluation for Code Review Bots
Explore the challenges and limitations of automated evaluation methods for code review bots in real-world software development settings.
AI Harms and Intersectionality: Insights from 5300 Reports
Explore how AI harms are amplified by intersectional identities and why risk assessments must consider multiple factors beyond race and gender.