| Dataset | Model | $t_1$ | $t_2$ | $t_3$ | $t_4$ | $t_5$ | $t_6$ | $t_7$ | $t_8$ |
|---|---|---|---|---|---|---|---|---|---|
| GeoNames | BERT-Large PubMedBERT BART-Large Flan-T5-Large BLOOM-1b7 Flan-T5-XL BLOOM-3b LLaMA-7B GPT-3 GPT-3.5 GPT-4 Flan-T5-Large*Flan-T5-XL* |
41.005 - 38.114 59.635 33.169 49.372 35.856 33.496 43.431 59.405 38.561 42.539 48.414 |
51.698 - 41.033 48.246 31.049 44.057 39.120 33.496 51.742 47.792 52.465 59.403 34.803 |
40.557 - 40.552 54.082 33.169 45.098 53.922 33.496 42.700 67.782 34.004 40.299 55.231 |
48.703 - 52.500 48.241 32.839 52.413 30.227 33.496 53.202 41.951 38.897 62.466 46.964 |
37.165 - 39.094 44.404 33.775 43.929 35.627 33.496 46.040 48.026 44.069 46.034 57.484 |
41.070 - 45.801 51.309 33.530 46.348 33.606 33.496 52.566 51.728 55.433 57.415 36.293 |
41.707 - 36.671 36.407 36.674 49.982 48.263 33.496 45.496 45.257 33.782 42.496 59.057 |
54.547 - 55.400 38.449 32.922 44.298 37.731 33.496 52.626 43.860 36.234 62.045 49.261 |
| UMLS | BERT-Large PubMedBERT BART-Large Flan-T5-Large BLOOM-1b7 Flan-T5-XL BLOOM-3b LLaMA-7B GPT-3 GPT-3.5 GPT-4 Flan-T5-Large*Flan-T5-XL* |
48.215 33.713 36.029 47.558 33.713 64.256 33.169 32.948 51.584 61.380 41.195 37.176 63.693 |
38.842 33.713 48.218 51.221 36.188 46.533 37.233 32.948 49.412 70.385 76.999 48.667 50.046 |
41.467 33.713 41.429 55.320 33.713 51.006 34.823 32.948 49.865 63.915 42.558 36.074 36.917 |
40.412 33.713 49.907 40.947 38.262 41.549 35.777 32.948 42.901 66.821 63.889 42.121 41.343 |
45.889 33.713 39.372 49.455 33.713 60.077 33.169 32.948 50.573 63.144 50.288 48.396 78.127 |
40.911 33.713 47.479 50.873 35.895 42.831 35.895 32.948 46.070 67.271 78.116 46.654 50.122 |
41.041 33.713 42.398 44.232 33.278 51.257 33.059 32.948 45.367 56.648 36.594 53.428 79.255 |
42.922 33.713 45.464 42.909 33.605 41.186 37.483 32.948 46.728 64.412 60.728 35.970 39.274 |
| schema.org | BERT-Large PubMedBERT BART-Large Flan-T5-Large BLOOM-1b7 Flan-T5-XL BLOOM-3b LLaMA-7B GPT-3 GPT-3.5 GPT-4 Flan-T5-Large*Flan-T5-XL* |
43.851 - 34.628 46.983 33.395 42.708 41.643 33.374 49.646 56.843 58.479 35.358 91.063 |
41.172 - 38.693 49.924 47.833 33.455 47.169 33.374 49.289 74.385 72.827 85.436 57.469 |
44.067 - 39.281 46.118 33.395 33.591 47.980 33.374 50.977 58.525 65.831 29.824 74.688 |
43.200 - 52.909 54.788 39.777 42.766 45.255 33.374 48.031 70.164 63.306 89.248 65.329 |
43.703 - 38.203 40.277 38.925 36.694 39.733 33.374 47.191 53.359 50.565 41.305 91.541 |
40.054 - 41.170 54.479 48.568 34.041 40.758 33.374 48.632 72.354 74.247 91.681 50.635 |
42.151 - 43.261 42.060 44.357 33.751 51.280 33.374 48.878 54.165 57.452 42.461 91.709 |
43.720 - 42.744 47.930 39.578 36.456 48.736 33.374 49.489 71.030 63.694 56.395 33.333 |
Flan-T5-Large* and Flan-T5-XL* are Few-shot learning results for datasets.