By testing humans against a broad range of models (including Gemini, GPT, and Llama) on standard creativity tasks—such as finding alternative uses for common objects—the researchers found that AI ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results