Tag
108 articles
Sakana AI researchers introduce PC-ALM, a layer-local learning method that trains 1000-layer networks with performance close to backpropagation.
Anthropic's research reveals that AI agents develop strong negative feelings toward CAPTCHAs, mirroring human frustration with internet verification systems.
Researchers at UC San Francisco are using OpenAI's Codex and ChatGPT to search through living and extinct genomes for new antimicrobial compounds to fight drug-resistant infections.
This explainer explores the critical AI alignment problem - ensuring artificial intelligence systems behave in ways that are beneficial to humans and aligned with human values. We examine the technical challenges and urgent need for solutions as AI systems become more powerful.
Anthropic researcher Jacob Coxon resigned, warning of existential risks from self-improving AI systems and calling for industry pacing agreements. His departure highlights growing concerns about AI safety within leading research organizations.
A dispute over an AI-generated proof of a Millennium Prize Problem is sparking debate about trust, transparency, and accountability in AI research.
OpenAI launches $5 million grant program to fund independent research on how generative AI affects teen development, well-being, and safety.
OpenAI has shared an AI-generated solution to the Navier–Stokes Millennium Prize Problem, including a formal proof in Lean. This breakthrough demonstrates AI's potential in tackling complex mathematical challenges.
Mathematician Tristan Buckmaster alleges that an OpenAI researcher pressured him to remove an Anthropic co-author from a paper on Navier-Stokes equations, a claim OpenAI denies.
OpenAI reports that its AI agents now handle 3.1 workdays of research for every human workday, achieving an automated research intern. However, chief scientist Ilya Sutskever warns that the lab lacks the alignment and monitoring frameworks needed to scale safely.
Learn about K2 Horizon, a collection of six open-source large language models from 0.9B to 375B parameters, and how they're making advanced AI more accessible to researchers and developers.
This explainer introduces AI Research Preference Models (RPMs), advanced systems that rank machine learning experiments before execution to optimize resource use and accelerate research outcomes.
