Update docs.md
Browse files
docs.md
CHANGED
|
@@ -1,10 +1,21 @@
|
|
| 1 |
<div style="display: flex; align-items: center; justify-content: space-between; width: 100%;">
|
| 2 |
-
<img
|
| 3 |
-
|
| 4 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
</div>
|
| 6 |
|
| 7 |
-
|
| 8 |
<h2>📜 Background</h2>
|
| 9 |
<p>Recent advances in <strong>Large Language Models (LLMs)</strong> have demonstrated transformative potential in improving healthcare delivery and clinical research. By combining extensive pretraining with supervised instruction tuning across diverse tasks, LLMs excel in natural language understanding, generation, and reasoning. These capabilities allow LLMs to serve as versatile, general-purpose medical assistants.</p>
|
| 10 |
<p>Despite this promise, concerns remain around the <strong>reliability and clinical validity</strong> of LLM-generated outputs. Real-world contexts often involve unstructured, multilingual text from <strong>electronic health records (EHRs)</strong>, and require support for tasks like phenotype identification and event extraction that remain underexplored. Moreover, the scarcity of <strong>multilingual benchmarks</strong> further limits the global applicability of LLMs in medicine.</p>
|
|
|
|
| 1 |
<div style="display: flex; align-items: center; justify-content: space-between; width: 100%;">
|
| 2 |
+
<img
|
| 3 |
+
src="https://cdn-uploads.huggingface.co/production/uploads/67a040fb6934f9aa1c866f99/ASGmOZ2FH1wWCE51QJrBZ.png"
|
| 4 |
+
alt="Leaderboard Image 1"
|
| 5 |
+
style="width: 30%; object-fit: contain;"
|
| 6 |
+
/>
|
| 7 |
+
<img
|
| 8 |
+
src="https://cdn-uploads.huggingface.co/production/uploads/67a040fb6934f9aa1c866f99/FsuzhBUMHOUVOV3raharV.png"
|
| 9 |
+
alt="Leaderboard Image 2"
|
| 10 |
+
style="width: 30%; object-fit: contain;"
|
| 11 |
+
/>
|
| 12 |
+
<img
|
| 13 |
+
src="https://cdn-uploads.huggingface.co/production/uploads/67a040fb6934f9aa1c866f99/xQqbGXh0y6zIV78Cw6Vpq.png"
|
| 14 |
+
alt="Leaderboard Image 3"
|
| 15 |
+
style="width: 30%; object-fit: contain;"
|
| 16 |
+
/>
|
| 17 |
</div>
|
| 18 |
|
|
|
|
| 19 |
<h2>📜 Background</h2>
|
| 20 |
<p>Recent advances in <strong>Large Language Models (LLMs)</strong> have demonstrated transformative potential in improving healthcare delivery and clinical research. By combining extensive pretraining with supervised instruction tuning across diverse tasks, LLMs excel in natural language understanding, generation, and reasoning. These capabilities allow LLMs to serve as versatile, general-purpose medical assistants.</p>
|
| 21 |
<p>Despite this promise, concerns remain around the <strong>reliability and clinical validity</strong> of LLM-generated outputs. Real-world contexts often involve unstructured, multilingual text from <strong>electronic health records (EHRs)</strong>, and require support for tasks like phenotype identification and event extraction that remain underexplored. Moreover, the scarcity of <strong>multilingual benchmarks</strong> further limits the global applicability of LLMs in medicine.</p>
|