1 paper
Robin Linzmayer, Georgianna Lin, Di Coneybeare +22
We introduce AcuityBench, a benchmark for evaluating whether language models identify the appropriate urgency of care from user medical presentations. Existing health benchmarks em…