VitalDB - AnesLLM: a dataset and benchmark based on real intraoperative anaesthetic decisions 1.0
| split | label | n | mean | std | min | p25 | p50 | p75 | max |
| dev | result_1 | 423 | 0.5369 | 0.1770 | 0.2204 | 0.3920 | 0.5154 | 0.6595 | 0.9973 |
| dev | result_aux | 423 | 0.2355 | 0.0933 | 0.0011 | 0.1783 | 0.2373 | 0.2881 | 0.4773 |
| train | result_1 | 80933 | 0.5707 | 0.1695 | 0.2026 | 0.4430 | 0.5549 | 0.6831 | 0.9993 |
| train | result_aux | 80933 | 0.2302 | 0.1052 | 0.0002 | 0.1551 | 0.2286 | 0.2998 | 0.4953 |
| test | result_1 | 3457 | 0.5705 | 0.1726 | 0.2250 | 0.4398 | 0.5562 | 0.6873 | 0.9975 |
| test | result_aux | 3457 | 0.2293 | 0.1031 | 0.0009 | 0.1559 | 0.2311 | 0.2961 | 0.4928 |
| overall | result_1 | 84813 | 0.5706 | 0.1697 | 0.2026 | 0.4425 | 0.5548 | 0.6831 | 0.9993 |
| overall | result_aux | 84813 | 0.2302 | 0.1051 | 0.0002 | 0.1552 | 0.2287 | 0.2996 | 0.4953 |