Microsoft Research presented with BioGPT a domain-specific generative model that was trained on biomedical literature and achieved best-in-class performance above the human level [paper, article].
Category: Medicine
In a recent paper from Dec 26, 2022, Google demonstrates that its large language model Med-PaLM, based on 540 billion parameters with a special instruction prompt tuning for the medical domain, reaches almost clinician’s performance on new medical benchmarks MultiMedQA (benchmark combining six existing open question answering datasets spanning professional medical exams, research, and consumer queries) and HealthSearchQA (a new free-response dataset of medical questions searched online). The evaluation of the answers considering factuality, precision, possible harm, and bias was done by human experts.
