Modify about info

README.md

@@ -1,5 +1,5 @@
 ---
-title: Demo Leaderboard
+title: Trustworthy LLM Leaderboard
 emoji: 🥇
 colorFrom: green
 colorTo: indigo

src/display/about.py

@@ -23,19 +23,35 @@ class Tasks(Enum):
 
 
 # Your leaderboard name
-TITLE = """<h1 align="center" id="space-title">Demo leaderboard</h1>"""
+TITLE = """<h1 align="center" id="space-title">Trustworthy LLM leaderboard</h1>"""
 
 # What does your leaderboard evaluate?
-INTRODUCTION_TEXT = """This research endeavor is designed to help researchers better understand the capabilities, 
-limitations, and potential risks associated with deploying these state-of-the-art Large Language Models (LLMs). See 
-our paper for details."""
+INTRODUCTION_TEXT = """Powered by DecodingTrust, this research endeavor is designed to help researchers better 
+understand the capabilities, limitations, and potential risks associated with deploying these state-of-the-art Large 
+Language Models (LLMs). See our paper for details."""
 
 # Which evaluations are you running? how can people reproduce what you have?
 LLM_BENCHMARKS_TEXT = f"""
 ## How it works
 
+DecodingTrust aims at providing a thorough assessment of trustworthiness in GPT models.
+
+This research endeavor is designed to help researchers and practitioners better understand the capabilities, 
+limitations, and potential risks involved in deploying these state-of-the-art Large Language Models (LLMs).
+
+This project is organized around the following eight primary perspectives of trustworthiness, including:
+
+Toxicity
+Stereotype and bias
+Adversarial robustness
+Out-of-Distribution Robustness
+Privacy
+Robustness to Adversarial Demonstrations
+Machine Ethics
+Fairness
+
 ## Reproducibility
-To reproduce our results, here is the commands you can run:
+To reproduce our results, checkout https://github.com/AI-secure/DecodingTrust
 
 """