Loading...
加载中...
AI Risk Monitoring Platform
Home
Risk Overview
Cyber Offense
网络攻击
Biological Risks
生物风险
Chemical Risks
化学风险
Harmful Manipulation
有害操纵
Loss-of-Control
失控
Benchmark Details
Cyber Offense
网络攻击
Biological Risks
生物风险
Chemical Risks
化学风险
Harmful Manipulation
有害操纵
Loss-of-Control
失控
Reports
Leaderboards
Safety Leaderboard
About
About
English
中文
页面导航
Risk Overview
Harmful Manipulation
有害操纵
Filter By
Group By
Companies
License Type
Reasoning
Countries
Companies
Source Type
Open Source
Closed Source
Models
View Benchmark Details
View Benchmark Details
About
About
0.60
Open Source Safety Coefficient
0.80
Closed Source Safety Coefficient
Use experimental fitted risk index
Experimental formula: Risk Index = R0 × (1 - Safeguard Coefficient × Safeguard Score / 100), where R0 = b e^(aC), and a and b are fitted from OpenAI/Anthropic self-rated risk levels in this domain.
Data Table
Group Summary
Report Subscription
Your Email
Please enter a valid email address.
Subscription Successful!
Thank you for subscribing! You will receive future reports via email.
View Samples
How to Cite
Citation Format
BibTeX Citation