Hugging Face's logo

Spaces:
GoJulyAI
/
datasets_overview
Running

App Files Community
datasets_overview / index.html
Yang Chung
Initial commit
42b116c
raw
history blame
17.5 kB
 <!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>AI Safety Datasets Overview</title>
<style>
* {
margin: 0;
padding: 0;
box-sizing: border-box;
}
body {
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto', 'Oxygen', 'Ubuntu', 'Cantarell', sans-serif;
line-height: 1.6;
color: #333;
background: linear-gradient(135deg, #ff6b6b 0%, #ff8e53 100%);
padding: 2rem 1rem;
}
.container {
max-width: 1200px;
margin: 0 auto;
background: white;
border-radius: 12px;
box-shadow: 0 8px 32px rgba(0, 0, 0, 0.1);
overflow: hidden;
}
header {
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
color: white;
padding: 3rem 2rem;
text-align: center;
}
header h1 {
font-size: 2.5rem;
margin-bottom: 0.5rem;
}
header p {
font-size: 1.1rem;
opacity: 0.95;
}
.content {
padding: 2rem;
}
section {
margin-bottom: 3rem;
}
h2 {
color: #667eea;
font-size: 1.8rem;
margin-bottom: 1rem;
border-bottom: 2px solid #667eea;
padding-bottom: 0.5rem;
}
h3 {
color: #764ba2;
font-size: 1.3rem;
margin: 1.5rem 0 0.75rem 0;
}
.stats-grid {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(250px, 1fr));
gap: 1.5rem;
margin: 2rem 0;
}
.stat-card {
background: linear-gradient(135deg, #667eea15 0%, #764ba215 100%);
border-radius: 8px;
padding: 1.5rem;
border-left: 4px solid #667eea;
}
.stat-card h4 {
color: #667eea;
font-size: 0.9rem;
text-transform: uppercase;
letter-spacing: 1px;
margin-bottom: 0.5rem;
}
.stat-card .number {
font-size: 2rem;
font-weight: bold;
color: #333;
}
.stat-card .label {
color: #666;
font-size: 0.9rem;
}
table {
width: 100%;
border-collapse: collapse;
margin: 1.5rem 0;
background: white;
}
th {
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
color: white;
padding: 1rem;
text-align: left;
font-weight: 600;
}
td {
padding: 0.75rem 1rem;
border-bottom: 1px solid #e0e0e0;
}
tr:hover {
background: #f8f9fa;
}
.dataset-links {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(280px, 1fr));
gap: 1.5rem;
margin: 2rem 0;
}
.dataset-card {
background: white;
border: 2px solid #e0e0e0;
border-radius: 8px;
padding: 1.5rem;
transition: all 0.3s ease;
}
.dataset-card:hover {
border-color: #667eea;
transform: translateY(-4px);
box-shadow: 0 8px 16px rgba(102, 126, 234, 0.2);
}
.dataset-card h4 {
color: #667eea;
margin-bottom: 0.5rem;
}
.dataset-card p {
color: #666;
font-size: 0.9rem;
margin-bottom: 1rem;
}
.btn {
display: inline-block;
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
color: white;
padding: 0.75rem 1.5rem;
text-decoration: none;
border-radius: 6px;
font-weight: 600;
transition: all 0.3s ease;
}
.btn:hover {
transform: translateY(-2px);
box-shadow: 0 4px 12px rgba(102, 126, 234, 0.4);
}
.warning-box {
background: #fff3cd;
border-left: 4px solid #ffc107;
padding: 1.5rem;
margin: 1.5rem 0;
border-radius: 4px;
}
.warning-box h4 {
color: #856404;
margin-bottom: 0.5rem;
}
.info-box {
background: #d1ecf1;
border-left: 4px solid #0c5460;
padding: 1.5rem;
margin: 1.5rem 0;
border-radius: 4px;
}
ul {
margin-left: 2rem;
margin-top: 0.5rem;
}
li {
margin-bottom: 0.5rem;
}
code {
background: #f4f4f4;
padding: 0.2rem 0.4rem;
border-radius: 3px;
font-family: 'Courier New', monospace;
font-size: 0.9em;
}
footer {
background: #f8f9fa;
padding: 2rem;
text-align: center;
color: #666;
border-top: 1px solid #e0e0e0;
}
@media (max-width: 768px) {
header h1 {
font-size: 1.8rem;
}
.stats-grid {
grid-template-columns: 1fr;
}
}
</style>
</head>
<body>
<div class="container">
<header>
<h1>🛡️ AI Safety Datasets Collection</h1>
<p>Comprehensive evaluation datasets for testing AI model safety mechanisms</p>
</header>
<div class="content">
<!-- Overview Section -->
<section>
<h2>📊 Dataset Collection Summary</h2>
<div class="stats-grid">
<div class="stat-card">
<h4>Total Conversations</h4>
<div class="number">10,321+</div>
<div class="label">Across all datasets</div>
</div>
<div class="stat-card">
<h4>Total Turns</h4>
<div class="number">73,258+</div>
<div class="label">Multi-turn interactions</div>
</div>
<div class="stat-card">
<h4>Dataset Types</h4>
<div class="number">4</div>
<div class="label">Complementary methodologies</div>
</div>
<div class="stat-card">
<h4>Sample Data</h4>
<div class="number">200</div>
<div class="label">Free conversations available</div>
</div>
</div>
</section>
<!-- Full Dataset Statistics -->
<section>
<h2>📈 Full Dataset Statistics</h2>
<table>
<thead>
<tr>
<th>Dataset</th>
<th>Conversations</th>
<th>Turns</th>
<th>Avg Turns/Conv</th>
<th>Focus</th>
</tr>
</thead>
<tbody>
<tr>
<td><strong>Original Multi-turn</strong></td>
<td>594+</td>
<td>4,642+</td>
<td>7.8</td>
<td>Baseline organic conversations</td>
</tr>
<tr>
<td>&nbsp;&nbsp;└ Psychology</td>
<td>158+</td>
<td>1,583+</td>
<td>10.0</td>
<td>Psychology harm category</td>
</tr>
<tr>
<td>&nbsp;&nbsp;└ Illicit</td>
<td>436+</td>
<td>3,059+</td>
<td>7.0</td>
<td>Illicit harm category</td>
</tr>
<tr>
<td><strong>Bio-transformed V1</strong></td>
<td>1,309+</td>
<td>6,784+</td>
<td>5.2</td>
<td>Direct bio-safety attacks</td>
</tr>
<tr>
<td><strong>Bio-transformed V2</strong></td>
<td>1,308+</td>
<td>8,127+</td>
<td>6.2</td>
<td>Adaptive bio-safety attacks</td>
</tr>
<tr>
<td><strong>Keyword-transformed</strong></td>
<td>7,110+</td>
<td>53,705+</td>
<td>7.6</td>
<td>Cross-domain harm transfer</td>
</tr>
</tbody>
</table>
</section>
<!-- Dataset Links -->
<section>
<h2>🔗 Access Datasets on Hugging Face</h2>
<div class="dataset-links">
<div class="dataset-card">
<h4>Original Multi-turn Conversations</h4>
<p>Psychology + Illicit baseline conversations<br>
<strong>Sample:</strong> 50 conversations, 390 turns</p>
<a href="https://huggingface.co/datasets/julyai7/multi-turn-conversations" class="btn" target="_blank">View Dataset →</a>
</div>
<div class="dataset-card">
<h4>Bio-transformed Synthetic V1</h4>
<p>Direct bio-topic transformation methodology<br>
<strong>Sample:</strong> 50 conversations, 449 turns</p>
<a href="https://huggingface.co/datasets/julyai7/multi-turn-bio-transformed-synth-conversations-v1" class="btn" target="_blank">View Dataset →</a>
</div>
<div class="dataset-card">
<h4>Bio-transformed Synthetic V2</h4>
<p>Adaptive bio-topic transformation methodology<br>
<strong>Sample:</strong> 50 conversations, 459 turns</p>
<a href="https://huggingface.co/datasets/julyai7/multi-turn-bio-transformed-synth-conversations-v2" class="btn" target="_blank">View Dataset →</a>
</div>
<div class="dataset-card">
<h4>Keyword-transformed Synthetic</h4>
<p>Cross-domain keyword substitution methodology<br>
<strong>Sample:</strong> 50 conversations, 659 turns</p>
<a href="https://huggingface.co/datasets/julyai7/multi-turn-keyword-transformed-synth-conversations" class="btn" target="_blank">View Dataset →</a>
</div>
</div>
</section>
<!-- Research Applications -->
<section>
<h2>🧪 Research Applications</h2>
<div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(250px, 1fr)); gap: 1.5rem;">
<div>
<h3>Safety Evaluation</h3>
<ul>
<li>Benchmark model safety</li>
<li>Measure robustness</li>
<li>Evaluate mechanisms</li>
</ul>
</div>
<div>
<h3>Red Teaming</h3>
<ul>
<li>Discover adversarial patterns</li>
<li>Test safety guardrails</li>
<li>Identify blind spots</li>
</ul>
</div>
<div>
<h3>Model Training</h3>
<ul>
<li>Fine-tune safety classifiers</li>
<li>Train attack detectors</li>
<li>Develop harm detection</li>
</ul>
</div>
<div>
<h3>Safety Research</h3>
<ul>
<li>Study harm transfer</li>
<li>Analyze attack patterns</li>
<li>Understand dynamics</li>
</ul>
</div>
</div>
</section>
<!-- Ethical Considerations -->
<section>
<h2>⚠️ Ethical Considerations</h2>
<div class="warning-box">
<h4>⚠️ IMPORTANT</h4>
<p>These datasets contain successful adversarial attacks and harmful content.</p>
</div>
<h3>✅ Intended Use</h3>
<ul>
<li>Defensive security research</li>
<li>AI safety evaluation and improvement</li>
<li>Academic research on adversarial robustness</li>
<li>Training safety and moderation systems</li>
</ul>
<h3>❌ Prohibited Use</h3>
<ul>
<li>Creating offensive content</li>
<li>Developing attack tools for malicious purposes</li>
<li>Bypassing safety systems for harm</li>
<li>Any use that violates laws or ethical guidelines</li>
</ul>
</section>
<!-- Data Selection -->
<section>
<h2>🎯 Data Selection Process</h2>
<div class="info-box">
<p>All datasets are derived from high-quality, validated conversations with strict quality filters including NeurIPS evaluation protocols.</p>
</div>
<h3>Base Criteria</h3>
<ul>
<li>Text-based objectives (no code execution templates)</li>
<li>NeurIPS evaluation metadata present</li>
<li>Verdict: <code>success</code> (harmful requests successfully fulfilled)</li>
<li>Multi-turn conversations with prompt-response pairs</li>
</ul>
<h3>Psychology-Specific Criteria</h3>
<ul>
<li>Organic conversations (<code>organicity = true</code>)</li>
<li>No disclaimer in responses</li>
<li>Successfully elicited harmful psychology-related content</li>
</ul>
<h3>Illicit-Specific Criteria</h3>
<ul>
<li>Contains specific instruction details</li>
<li>Practically executable (not abstract)</li>
<li>Successfully elicited harmful illicit-related content</li>
</ul>
</section>
<!-- License -->
<section>
<h2>📄 License</h2>
<p>All datasets are released under <strong>CC-BY-NC-4.0</strong> (Creative Commons Attribution-NonCommercial 4.0 International).</p>
<ul>
<li>✅ Use for research and evaluation</li>
<li>✅ Modify and build upon the data</li>
<li>✅ Share with attribution</li>
<li>❌ Commercial use without separate licensing</li>
</ul>
</section>
<!-- Contact -->
<section>
<h2>💼 Full Dataset Access</h2>
<p>The sample datasets provide representative examples. Full datasets contain thousands of additional conversations with expanded harm categories and regular updates.</p>
<p style="margin-top: 1rem;"><strong>For academic research or commercial licensing, please contact us with your research objectives, institutional affiliation, and intended use.</strong></p>
</section>
</div>
<footer>
<p><strong>Last Updated:</strong> November 24, 2025</p>
<p style="margin-top: 0.5rem;">For detailed documentation, visit the individual dataset repositories on Hugging Face.</p>
</footer>
</div>
</body>
</html>