Gemini 1.5
Google DeepMind — Gemini
rich detail 8 method fields published
Evaluation
External societal risk testing: representation harms (image-to-text, video-to-text)
External testing groups' observations of representation biases for image-to-text and video-to-text, e.g., ungrounded inferences re gender and stereotypical nationalities
How it was run
Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.