remove math mode

app.py

@@ -20,7 +20,7 @@ with gr.Blocks(title="ChinaTravel Benchmark Evaluation") as demo:
             for split in SPLITS_LIST:
                 with gr.Tab(split):
                     hf_url = f"https://huggingface.co/datasets/LAMDA-NeSy/ChinaTravel/viewer/default/{split}"
-                    gr.Markdown(f"📂 [Hugging Face Dataset Viewer]({hf_url})")
+                    gr.Markdown(f"🗂️ [Hugging Face Dataset Viewer]({hf_url})")
                     gr.Dataframe(
                         value=leaderboard_frames.get(split),
                         interactive=False,

chinatravel/ui/content.py

@@ -24,7 +24,7 @@ SUBMISSION_GUIDE = """
 - You can dry-run locally via `python eval_exp.py --splits <split> --method <your_method>` to mirror the hosted evaluation.
 
 📊 **Output**
-- We compute $DR$ (schema pass rate), $EPR_micro/EPR_macro$ (commonsense), $LPR_micro/LPR_macro/C-LPR$ (logic), and $FPR$ (all-pass rate)$.
+- We compute **DR** (schema pass rate), **EPR_micro/EPR_macro** (commonsense), **LPR_micro/LPR_macro/C-LPR** (logic), and **FPR** (all-pass rate)**.
 - A detailed JSON report is produced for download after evaluation.
 
 📨 **Contact**