update the formatting. and links

_header.md

@@ -2,4 +2,4 @@
 
 # 🦓 ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning
 <!-- [📑 FnF Paper](https://arxiv.org/abs/2305.18654) |  -->
-[📰 Blog](https://huggingface.co/blog/yuchenlin/zebra-logic) [💻 GitHub](https://github.com/WildEval/ZeroEval) | [🤗 HuggingFace](https://huggingface.co/collections/allenai/zebra-logic-bench-6697137cbaad0b91e635e7b0) | [🐦 X](https://twitter.com/billyuchenlin/) | [💬 Discussion](https://huggingface.co/spaces/allenai/ZebraLogicBench-Leaderboard/discussions) | Updated: **{LAST_UPDATED}**
+[📑 Paper](https://arxiv.org/abs/2502.01100) | [💻 GitHub](https://github.com/WildEval/ZeroEval) | [🤗 Dataset](https://huggingface.co/datasets/WildEval/ZebraLogic) | Updated: **{LAST_UPDATED}**

app.py

@@ -190,7 +190,7 @@ def data_load(result_file):
     main_ordered_columns = ORDERED_COLUMN_NAMES
     # filter the data with Total Puzzles == 1000
 
-    click_url = True
+    click_url = False
     # read json file from the result_file
     with open(result_file, "r") as f:
         raw_data = json.load(f)
@@ -203,7 +203,8 @@ def data_load(result_file):
                 pass
     original_df = pd.DataFrame(raw_data)
     original_df = original_df[original_df["Total Puzzles"] == 1000]
-
+    # let's remove some models in the list: ["o1-preview-2024-09-12-v2", "o1-mini-2024-09-12-v2", "o1-mini-2024-09-12"]
+    original_df = original_df[~original_df["Model"].isin(["o1-preview-2024-09-12-v2", "o1-mini-2024-09-12-v2", "o1-mini-2024-09-12"])]
     original_df = post_processing(original_df, column_names_main, ordered_columns=main_ordered_columns, click_url=click_url, rank_column=RANKING_COLUMN)
     print(f"original_df.columns: {original_df.columns}")
     # print(original_df.columns)