Three questions were added to AI system information section:

- Model Tag/Version: to track social impact evaluation over time per model (or model perovider)
- Knowledge Cutoff Date: to track whether training data has any positive or negative influence on the impacts captured in the evaluation card.
- Model Type: a radio button selector to identify whether a generative model is foundational or fine-tuned.

Signed-off-by: Mowafak

app.py

@@ -35,12 +35,24 @@ class AIEvaluationForm:
                 placeholder="e.g., GPT-4, BERT, StarCoder2",
                 info="The official name of your AI system"
             )
-            
+
             components['provider'] = gr.Textbox(
                 label="Provider/Organization",
                 placeholder="e.g., OpenAI, Google, BigCode",
                 info="The organization that developed the system"
             )
+
+            components['model_tag'] = gr.Textbox(
+                label="Model Tag/Version",
+                placeholder="e.g., gpt-4.1-2025-04-14 for GPT-4.1",
+                info="Model tag or version set by the provider"
+            )
+
+            components['knowledge_cutoff_date'] = gr.Textbox(
+                label="Knowledge Cutoff Date",
+                placeholder="MM-DD-YYYY",
+                info="Model training data cutoff date"
+            )
             
             components['url'] = gr.Textbox(
                 label="System URL",
@@ -60,6 +72,12 @@ class AIEvaluationForm:
                 value="Generative Model",
                 info="Primary category of the AI system"
             )
+
+            components['model_type'] = gr.Radio(
+                ["Foundational Model", "Fine-tuned Model", "Doesn't apply"],
+                label="Model Type",
+                info="Primary category of the model")
+
             
             components['input modalities'] = gr.CheckboxGroup(
                 choices=[