Upload 2 files

app.py

@@ -111,7 +111,7 @@ with gr.Blocks(theme=theme) as app:
         with gr.TabItem("Report Generation"):
             with gr.Row():
                 with gr.Column(scale=7):
-                    gr.Markdown("Report Generation Leaderboard: LLMs generate reports with just the prompt, which are then evaluated by gpt-oss-120b (mixed) judge with the lite dataset (160 samples) \nEvaluation and cost estimation last performed on 5 Dec 2025.")
+                    gr.Markdown("Report Generation Leaderboard: LLMs generate reports with just the prompt, which are then evaluated by gpt-oss-120b (mixed) judge with the lite dataset (160 samples) \nEvaluation and cost estimation last performed on 11 Dec 2025.")
                 
             with gr.Tabs(elem_id="inner-tabs", elem_classes="tabs-small") as tabs:
                 with gr.TabItem("Leaderboard"):

report_generation.jsonl

@@ -44,3 +44,4 @@
 {"Model": "Anthropic/claude-opus-4.5 (Thinking)", "Category": "Closed-source Reasoning", "Overall": 54.2, "Physics": 40.5, "Chemistry": 68.5, "Finance": 39.0, "Consulting": 68.7, "Extraction": 48.6, "Reasoning": 56.1, "Style": 65.7, "Response Characters": 9400, "Input Tokens": 560, "Output Tokens": 17627, "Cost": 70.96}
 {"Model": "DeepSeek-AI/DeepSeek-V3.2 (Thinking)", "Category": "Open-weight Reasoning", "Overall": 45.7, "Physics": 30.0, "Chemistry": 55.8, "Finance": 33.0, "Consulting": 64.1, "Extraction": 39.1, "Reasoning": 45.9, "Style": 58.2, "Response Characters": 4346, "Input Tokens": 456, "Output Tokens": 13245, "Cost": 0.87}
 {"Model": "DeepSeek-AI/DeepSeek-V3.2", "Category": "Open-weight Instruct", "Overall": 50.5, "Physics": 42.1, "Chemistry": 57.8, "Finance": 37.1, "Consulting": 65.0, "Extraction": 43.3, "Reasoning": 51.7, "Style": 53.1, "Response Characters": 8266, "Input Tokens": 456, "Output Tokens": 2755, "Cost": 0.2}
+{"Model": "OpenAI/GPT-5.2 (xhigh)", "Category": "Closed-source Reasoning", "Overall": 55.1, "Physics": 42.7, "Chemistry": 72.6, "Finance": 35.6, "Consulting": 69.4, "Extraction": 47.3, "Reasoning": 57.8, "Style": 75.7, "Response Characters": 7010, "Input Tokens": 3191, "Output Tokens": 22965, "Cost": 52.34}