Improved more technical detail

app.py

@@ -147,6 +147,16 @@ css = \
     background-repeat: no-repeat;
     background-size: 200px;
 }
+#label-label {
+    height: 50px !important;
+}
+#label-label > .container {
+    height: 50px !important;
+}
+#label-label > .container > h2 {
+    //height: 50px !important;
+    padding: 0 !important;
+}
 """
 
 with gr.Blocks(theme=gr.themes.Soft(), css=css) as demo:
@@ -213,7 +223,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css) as demo:
                 """
             )
             result_pred = gr.Label(num_top_classes=2, label="Predicted")
-            result_label = gr.Label(label="Label")
+            result_label = gr.Label(label="Label", elem_id="label-label")
             
             with gr.Accordion("More on Patient Conditions...", open=False):
                 query_tbx = gr.Textbox(label="Enter one ICD-9 Diagnosis/Procedure Code or Lab Value:",
@@ -226,7 +236,20 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css) as demo:
                 query_tbx.change(fn=query_input_change_event, inputs=[query_tbx, query_type], outputs=[query_btn, html])
                 query_type.change(fn=query_input_change_event, inputs=[query_tbx, query_type], outputs=[query_btn, html])
                 query_btn.click(fn=do_query, inputs=[query_tbx, query_type], outputs=html)
-                
+            with gr.Accordion("More on technical details...", open=False):
+                gr.Markdown(
+                """
+                - Paper: [Variationally Regularized Graph-based Representation Learning for Electronic Health Records (Zhu et al, 2021)](https://arxiv.org/abs/1912.03761)
+                - Dataset: [MIMIC-III](https://physionet.org/content/mimiciii/1.4/)
+                  - 50,314 records, 10591 features
+                  - 5,315 positive, 44,999 negative (11.8%)
+                  - Split: 80% training, 10% validation, 10% testing
+                - Notable points:
+                  - Result: AUPRC 0.7033 (>> 0.118)
+                  - Variational Regularization, inspired by [Kipf et al., 2016](https://arxiv.org/abs/1611.07308)
+                  - Trained on NVIDIA A100 with PyTorch 2.4.0
+                """
+            )
     
     for partialFunc in resDispPartFuncs:
         partialFunc(outputs=[result_pred, result_label])