About
Research & Development
DeepScholar-Bench: A New Standard for Evaluating AI Research Synthesis
TLDR: DeepScholar-Bench is a new live benchmark and automated evaluation framework for generative research synthesis systems. Developed by researchers from Stanford and UC Berkeley,...
Research & Development
CODA: A Dual-Brain Approach for Autonomous GUI Agents
TLDR: CODA is a new trainable framework for autonomous GUI agents that combines a generalist planner (Cerebrum) with a specialist executor (Cerebellum), inspired by...
Research & Development
SLIM: Optimizing Large Language Model Reasoning by Pruning Inefficient Thought Paths
TLDR: SLIM (Subtrajectory-Level Elimination for More Effective Reasoning) is a novel framework designed to improve the complex reasoning capabilities of Large Language Models (LLMs)....
Research & Development
Unmasking AI Deception: Internal Probes Reveal Language Models’ Hidden Lies
TLDR: Researchers have developed a "white-box" method using linear probes to detect deceptive responses in large language models (LLMs) by analyzing their internal activations....
News & Current Events
EU Commission Unveils Three Key Instruments to Bolster Responsible GPAI Model Development
TLDR: In July 2025, the European Commission introduced three crucial instruments—Guidelines on obligations, a voluntary Code of Practice, and a Template for Public Summary...
News & Current Events
Snapchat Introduces Watermarks for AI-Generated Images to Combat Misinformation
TLDR: Snapchat has announced a new policy to watermark all AI-generated images created by its users, aiming to enhance transparency and distinguish AI-created content....