Update exp2-main.py

exp2-main.py

@@ -208,26 +208,65 @@ def generate_rating_prompt(section_title: str) -> str:
 def instructions_1():
     st.title("Experiment 2: LLM Scene Abstraction Evaluation")
     st.header("📖 Instructions (1/2)")
-    st.write(f"""
-    Welcome to Experiment 2! Here’s how it works:
+    # st.write(f"""
+    # Welcome to Experiment 2! Here’s how it works:
 
-    - You will read a sentence that contains a specific <span style='color:{HIGHLIGHT_COLOR}; font-weight:500;'>**keyword**</span>.
-    - You will then see <span style='color:{HIGHLIGHT_COLOR}; font-weight:500;'>**scene-level information about the keyword** in the given situation</span>, generated by a large language model (LLM).
+    # - You will read a sentence that contains a specific <span style='color:{HIGHLIGHT_COLOR}; font-weight:500;'>**keyword**</span>.
+    # - You will then see <span style='color:{HIGHLIGHT_COLOR}; font-weight:500;'>**scene-level information about the keyword** in the given situation</span>, generated by a large language model (LLM).
 
-        - The information is organized into three sections:
-          1. **Engaged Events** — What is happening to the keyword in this situation?
-          2. **Generalizable Properties** — What context-relevant properties of the keyword are revealed through this situation?
-          3. **Evoked Emotions** — What emotions are associated with the keyword in this scene, and why?
-    <br>
-    Your task is to **evaluate each section** based on how well it reflects the information conveyed in the original sentence.
+    #     - The information is organized into three sections:
+    #       1. **Engaged Events** — What is happening to the keyword in this situation?
+    #       2. **Generalizable Properties** — What context-relevant properties of the keyword are revealed through this situation?
+    #       3. **Evoked Emotions** — What emotions are associated with the keyword in this scene, and why?
+    # <br>
+    # Your task is to **evaluate each section** based on how well it reflects the information conveyed in the original sentence.
+
+    # - For each section, please rate the following dimensions on a 1–5 scale:
+    #   - **Accuracy** — How accurate is it? Is the content factually consistent with the sentence?
+    #   - **Completeness** — How complete and rich is it? Does it fully capture the relevant aspects of the keyword?
+    #   - **Interpretability** — How interpretable is it? Is it easy to understand?
 
-    - For each section, please rate the following dimensions on a 1–5 scale:
-      - **Accuracy** — How accurate is it? Is the content factually consistent with the sentence?
-      - **Completeness** — How complete and rich is it? Does it fully capture the relevant aspects of the keyword?
-      - **Interpretability** — How interpretable is it? Is it easy to understand?
+    # <br>
+    # If you have questions or feedback, please feel free to let us know via email.
+    # <br><br>
+
+    # """, unsafe_allow_html=True)
+    st.write(f"""
+    <p style='font-size:18px;'>
+    Welcome to Experiment 2! Here’s how it works:
+    </p>
+
+    <ul style='font-size:18px;'>
+        <li>You will read a sentence that contains a specific <span style='color:{HIGHLIGHT_COLOR}; font-weight:500;'><b>keyword</b></span>.</li>
+        <li>You will then see <span style='color:{HIGHLIGHT_COLOR}; font-weight:500;'><b>scene-level information about the keyword</b> in the given situation</span>, generated by a large language model (LLM).</li>
+    </ul>
+
+    <p style='font-size:18px;'>
+    The information is organized into three sections:
+    </p>
+    <ul style='font-size:18px;'>
+        <li><b>Engaged Events</b> — What is happening to the keyword in this situation?</li>
+        <li><b>Generalizable Properties</b> — What context-relevant properties of the keyword are revealed through this situation?</li>
+        <li><b>Evoked Emotions</b> — What emotions are associated with the keyword in this scene, and why?</li>
+    </ul>
+    <br>
+    <p style='font-size:18px;'>
+    Your task is to <b>evaluate each section</b> based on how well it reflects the information conveyed in the original sentence.
+    </p>
+
+    <p style='font-size:18px;'>
+    For each section, please rate the following dimensions on a 1–5 scale:
+    </p>
+    <ul style='font-size:18px;'>
+        <li><b>Accuracy</b> — How accurate is it? Is the content factually consistent with the sentence?</li>
+        <li><b>Completeness</b> — How complete and rich is it? Does it fully capture the relevant aspects of the keyword?</li>
+        <li><b>Interpretability</b> — How interpretable is it? Is it easy to understand?</li>
+    </ul>
 
     <br>
+    <p style='font-size:18px;'>
     If you have questions or feedback, please feel free to let us know via email.
+    </p>
     <br><br>
 
     """, unsafe_allow_html=True)