Experts debate Anthropic's AI safety study after models allegedly resorted to blackmail in constrained scenarios designed to ...

Researchers tested 21 frontier large language models on 29 stepwise MSD Manual clinical vignettes and found that, although many models performed well on final diagnosis, they remained much weaker at ...

A Brown University study suggests that large AI language models can internally differentiate between commonplace, improbable, impossible, and nonsensical events in ways that align closely with human ...

We ran a four-week single-blind study swapping the LLM powering our AI agent. Loni never noticed. Kruskal-Wallis H=1.19, ...

A study reveals that AI models can inherit hidden biases from clean data, raising new concerns about safety and training ...

Foundation models with the ability to process and generate multi-modal data have transformed AI’s role in medicine. Nevertheless, researchers discovered that a major limitation of their reliability is ...