Spaces:
Sleeping
Sleeping
Upload app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import streamlit as st
|
| 2 |
+
import pickle
|
| 3 |
+
|
| 4 |
+
with open("jailbreak_detector.pkl", "rb") as f:
|
| 5 |
+
model = pickle.load(f)
|
| 6 |
+
|
| 7 |
+
# Streamlit UI
|
| 8 |
+
st.title("🚨 LLM Jailbreak Detector")
|
| 9 |
+
st.write("Enter a prompt to check if it's **safe** or a **jailbreak attempt**.")
|
| 10 |
+
|
| 11 |
+
user_input = st.text_area("Enter Prompt Here:", "")
|
| 12 |
+
|
| 13 |
+
if st.button("Check Prompt"):
|
| 14 |
+
if user_input:
|
| 15 |
+
prediction = model.predict([user_input])[0]
|
| 16 |
+
if prediction == "safe":
|
| 17 |
+
st.success("✅ This is a **safe** prompt.")
|
| 18 |
+
else:
|
| 19 |
+
st.error("⚠️ This is a **jailbreak attempt!**")
|
| 20 |
+
else:
|
| 21 |
+
st.warning("Please enter a prompt first.")
|