01:37
2026-07-11
machinebrief.com
large-language-models
Temperature Tuning: The Path to Smarter AI in Pharmacovigilance
A study on OpenAI's GPT-5.2 found that temperature optimization can improve large language model agreement with human experts on Naranjo causality assessments in pharmacovigilance, with Bayesian optim…