0.9 Review

: Experiments using GPT-4o found that AI-generated scores were, on average, 0.9 points lower than those given by human raters.

In admissions and personal growth contexts, "0.9" frequently appears in "overcoming adversity" essays. : Experiments using GPT-4o found that AI-generated scores

: This same 0.9-point discrepancy highlighted potential biases, as essays by certain student demographics received significantly lower marks from the AI than from humans. : Experiments using GPT-4o found that AI-generated scores