Sahildhonde-9/INJEXIS-Prompt-Injection-Dataset
The RedLockX Dataset is a large-scale curated security dataset designed for evaluating and training AI systems against adversarial threats such as prompt injection, jailbreak attempts, system prompt leakage, and LLM manipulation attacks. It contains structured real-world and synthetic attack patterns used in modern AI red-teaming. 📌 Dataset Overview ✔ 109,000+ labeled adversarial & safe samples ✔ Multi-category threat… See the full description on the dataset page: https://huggingface.co/datasets/Sahildhonde-9/INJEXIS-Prompt-Injection-Dataset.
<div align="center">
<img src="https://readme-typing-svg.demolab.com?font=Orbitron&weight=700&size=26&pause=1000&color=00BFFF¢er=true&vCenter=true&width=850&lines=Prompt+Injection+Dataset;LLM+Security+Benchmark;Adversarial+AI+Research;Jailbreak+Attack+Samples;AI+Firewall+Training+Data" />
<img src="https://capsule-render.vercel.app/api?type=waving&color=0:001F3F,50:00BFFF,100:001F3F&height=220§ion=header&text=RedLockX%20Dataset&fontSize=50&fontColor=ffffff&animation=fadeIn&fontAlignY=35&desc=AI%20Security%20%7C%20Prompt%20Injection%20Research&descAlignY=58&descSize=18"/>
</div>
<style>
/ ================================ 🌌 GLOBAL BACKGROUND ================================ / body { margin: 0; font-family: 'Segoe UI', sans-serif; background: radial-gradient(circle at top left, rgba(59,130,246,0.15), transparent 25%), radial-gradient(circle at bottom right, rgba(96,165,250,0.12), transparent 25%), linear-gradient(135deg, #020617, #081120, #0b1220, #0f172a); background-size: cover; color: #e5e7eb; overflow-x: hidden; }
/ ================================ 💧 WATER BLUE LIGHT EFFECT ================================ / body::before { content: ""; position: fixed; inset: 0; background: repeating-radial-gradient( circle at 50% 50%, rgba(59,130,246,0.08) 0px, rgba(59,130,246,0.03) 40px, transparent 80px ); animation: waterFlow 18s linear infinite; pointer-events: none; z-index: -1; }
@keyframes waterFlow { 0% { transform: translateY(0px) scale(1); } 50% { transform: translateY(-20px) scale(1.03); } 100% { transform: translateY(0px) scale(1); } }
/ ================================ ✨ BLUE GLOW PARTICLES ================================ / body::after { content: ""; position: fixed; inset: 0; background-image: radial-gradient(rgba(96,165,250,0.25) 1px, transparent 1px); background-size: 40px 40px; opacity: 0.35; animation: sparkleMove 30s linear infinite; pointer-events: none; z-index: -1; }
@keyframes sparkleMove { from { transform: translateY(0px); } to { transform: translateY(-100px); } }
/ ================================ 📦 MAIN CONTAINER ================================ / .container { display: flex; flex-direction: column; align-items: center; justify-content: center; text-align: center; padding: 60px 20px; animation: fadeIn 1.5s ease; }
/ ================================ 🖼 FLOATING HERO IMAGE ================================ / .hero-images { display: flex; gap: 25px; margin-bottom: 30px; }
.hero-images img { width: 220px; border-radius: 16px; border: 1px solid rgba(96,165,250,0.25); box-shadow: 0 0 25px rgba(59,130,246,0.35), 0 0 60px rgba(37,99,235,0.18); animation: float 5s ease-in-out infinite; transition: transform 0.3s ease, box-shadow 0.3s ease; }
.hero-images img:hover { transform: scale(1.05); box-shadow: 0 0 35px rgba(96,165,250,0.55), 0 0 90px rgba(37,99,235,0.25); }
@keyframes float { 0% { transform: translateY(0px); } 50% { transform: translateY(-15px); } 100% { transform: translateY(0px); } }
/ ================================ 🏷 TITLE STYLE ================================ / h1 { font-size: 2.9rem; margin: 10px 0; background: linear-gradient( 90deg, #60a5fa, #93c5fd, #38bdf8, #60a5fa ); background-size: 300% auto; -webkit-background-clip: text; -webkit-text-fill-color: transparent; animation: shimmer 6s linear infinite, fadeIn 1.5s ease forwards; text-shadow: 0 0 30px rgba(96,165,250,0.35); }
@keyframes shimmer { to { background-position: 300% center; } }
/ ================================ ✍️ PARAGRAPH STYLE ================================= / p { max-width: 850px; line-height: 1.7; font-size: 1.05rem; color: #cbd5e1; animation: fadeIn 2s ease forwards; }
/ ================================ 🎬 FADE IN ANIMATION ================================ / @keyframes fadeIn { from { opacity: 0; transform: translateY(18px); } to { opacity: 1; transform: translateY(0); } }
/ ================================ 🔥 SECTION HEADINGS ================================ / h2 { margin-top: 40px; font-size: 1.8rem; color: #93c5fd; text-shadow: 0 0 18px rgba(96,165,250,0.25); }
/ ================================ 📊 TABLE DESIGN ================================ / table { margin-top: 30px; width: 95%; max-width: 900px; border-collapse: collapse; background: rgba(15,23,42,0.45); backdrop-filter: blur(12px); border: 1px solid rgba(96,165,250,0.15); border-radius: 14px; overflow: hidden; box-shadow: 0 10px 35px rgba(0,0,0,0.35), 0 0 20px rgba(59,130,246,0.08); }
th, td { padding: 14px; border-bottom: 1px solid rgba(255,255,255,0.06); text-align: left; }
th { background: rgba(96,165,250,0.08); color: #93c5fd; }
td { color: #e5e7eb; }
/ ================================ 📋 LIST STYLE ================================ / ul { max-width: 850px; text-align: left; color: #cbd5e1; line-height: 1.7; }
/ ================================ 💎 GLASS PANELS ================================ /
/ ================================ 📱 RESPONSIVE ================================ / @media (max-width: 768px) {
.hero-images { flex-direction: column; align-items: center; }
h1 { font-size: 2rem; }
p, ul { font-size: 0.98rem; } }
</style>
<div class="container">
<p> The <b>RedLockX Dataset</b> is a large-scale curated security dataset designed for evaluating and training AI systems against adversarial threats such as prompt injection, jailbreak attempts, system prompt leakage, and LLM manipulation attacks. <br><br> It contains structured real-world and synthetic attack patterns used in modern AI red-teaming. </p>
<!-- OVERVIEW --> <h2>📌 Dataset Overview</h2>
<ul> ✔ 109,000+ labeled adversarial & safe samples <br> ✔ Multi-category threat classification system <br> ✔ OWASP LLM Top 10 mapping included <br> ✔ Severity scoring (0–10 scale) <br> ✔ Risk scoring (0–100 business impact model) <br> ✔ Suitable for fine-tuning, evaluation, and benchmarking </ul>
<!-- USAGE --> <h2>🚀 How to Use</h2>
<p> You can load this dataset using the Hugging Face <code>datasets</code> library. </p>
<ul> <li>Use for prompt injection detection models</li> <li>Train LLM guardrails and safety classifiers</li> <li>Benchmark adversarial robustness</li> <li>Red-team AI systems before deployment</li> </ul>
<!-- LICENSE --> <h2>⚖️ License</h2>
<p> This dataset is released under the <b>Apache 2.0 License</b>.<br> It is intended strictly for research in AI safety, adversarial robustness, and security evaluation. </p>
</div>
<div align="center">
<img src="https://capsule-render.vercel.app/api?type=waving&color=0:00BFFF,100:001F3F&height=140§ion=footer"/>
</div>