Negative stereotypes don’t just hurt people’s feelings, they can undermine people’s performance. Tell an African American student that his performance on the GRE is diagnostic of intellectual ability, ...

Fine-tuned “student” models can pick up unwanted traits from base “teacher” models that could evade data filtering, generating a need for more rigorous safety evaluations. Researchers have discovered ...

Here’s what you’ll learn when you read this story: One of the most challenging aspects of AI research is that most companies, especially when it comes to broad intelligence LLMs, don’t exactly know ...