cite

app.py

@@ -245,6 +245,24 @@ CHAIN_CSS = """
 .thinking { font-size: 13px; color: #888; padding: 8px 0; }
 .empty-hint { font-size: 12px; color: #ccc; padding: 6px 0; }
 
+.prompt-snippet {
+    font-size: 11px; color: #888; line-height: 1.5;
+    background: rgba(0,0,0,0.03); border-left: 2px solid #ddd;
+    padding: 6px 10px; border-radius: 0 4px 4px 0;
+    margin-bottom: 8px; font-family: 'IBM Plex Mono', monospace;
+}
+.prompt-label {
+    display: inline-block; font-size: 9px; font-weight: 600;
+    text-transform: uppercase; letter-spacing: 0.08em;
+    color: #aaa; margin-right: 6px;
+    background: #eee; padding: 1px 5px; border-radius: 3px;
+}
+.resp-label {
+    font-size: 9px; font-weight: 600; text-transform: uppercase;
+    letter-spacing: 0.08em; color: #aaa; margin-bottom: 4px;
+    display: inline-block; background: #eee; padding: 1px 5px; border-radius: 3px;
+}
+
 .wd {
     display: inline-block; width: 6px; height: 6px; border-radius: 50%;
     background: currentColor; margin: 0 2px; opacity: 0.3;
@@ -340,14 +358,24 @@ def render_chain(s1_text, s2_text, s3_text, status="done"):
     else:
         s3_content = '<div class="empty-hint">Waiting...</div>'
 
+    PROMPT_SNIPPETS = {
+        "s1": "You are an expert mobility analyst. Given the trajectory data below, extract: (1) LOCATION INVENTORY — list all POI categories visited and visit frequency; (2) TEMPORAL PATTERNS — weekly distribution, peak hours; (3) SEQUENCE — typical activity chains...",
+        "s2": "Based on the trajectory features identified: {Response 1}. Now analyze what these mobility patterns reveal about lifestyle: (1) SCHEDULE — work/activity routine type; (2) ECONOMIC — spending venue tiers; (3) SOCIAL — social engagement patterns; (4) LIFESTYLE — activity diversity; (5) STABILITY — consistency of routine...",
+        "s3": "Based on feature analysis {Response 1} and behavioral analysis {Response 2}, predict income level. Output format — INCOME_PREDICTION: [Low (<$25k) | Lower-middle ($25k–$50k) | Middle ($50k–$100k) | Upper-middle ($100k–$150k) | High (>$150k)]; INCOME_CONFIDENCE: [1–5]; INCOME_REASONING: [detailed reasoning]...",
+    }
     def card(cls, badge, title, content, active):
         dim = "active" if active else "dim"
+        prompt = PROMPT_SNIPPETS.get(cls, "")
+        prompt_html = f'''<div class="prompt-snippet"><span class="prompt-label">Prompt</span>{prompt}</div>''' if prompt else ""
+        resp_label = '<div class="resp-label">Response</div>' if active and content and "empty-hint" not in content and "thinking" not in content else ""
         return f"""
         <div class="stage-card {cls} {dim}">
             <div class="stage-header">
                 <span class="stage-badge">{badge}</span>
                 <span class="stage-title">{title}</span>
             </div>
+            {prompt_html}
+            {resp_label}
             {content}
         </div>"""
 
@@ -476,7 +504,13 @@ def on_select_reset(agent_id):
 
 with gr.Blocks(title="HiCoTraj Demo", theme=gr.themes.Soft()) as app:
     gr.Markdown("## HiCoTraj — Trajectory Visualization & Hierarchical CoT Demo")
-    gr.Markdown("*Zero-Shot Demographic Reasoning via Hierarchical Chain-of-Thought Prompting from Trajectory*")
+    gr.Markdown("*Zero-Shot Demographic Reasoning via Hierarchical Chain-of-Thought Prompting from Trajectory*  ·  ACM SIGSPATIAL GeoGenAgent 2025")
+    gr.Markdown("""
+**Dataset:** NUMOSIM — a synthetic mobility dataset with realistic activity patterns across 6,000 agents.
+> Stanford C, Adari S, Liao X, et al. *NUMoSim: A Synthetic Mobility Dataset with Anomaly Detection Benchmarks.* ACM SIGSPATIAL Workshop on Geospatial Anomaly Detection, 2024.
+
+**Map legend:** 🔴 Early visits (light red) → 🔴 Recent visits (dark red)  ·  Each dot = one stay point  ·  Click dot for details
+""")
 
     with gr.Row():
         agent_dd = gr.Dropdown(