Unverified50% confidenceBenchmarkExact time
The RealToxicityPrompts dataset contains more than 100,000 prompts sampled from web text, used to quantify models' tendency to generate toxic content.
1
Sources
50%
Confidence
Long-term
Relevance
9/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedSimulated tweet tools carry the risk of being misused to create fake screenshots for misleading distribution.58% similarUnverifiedThe total amount of high-quality text data available on the internet is estimated by mainstream sources at approximately 10^13 to 10^14 tokens.55% similarUnverifiedDoubleVerify在2024年发布的报告显示，AI爬虫流量已占部分内容网站总流量的30%以上，但其中大量请求未正确标识来源身份55% similarUnverifiedAn internal study commissioned by Twitter in 2021 found that the platform's algorithm systematically amplified politically right-wing content and emotionally charged news.53% similarUnverified普通网页中真正有价值的信息可能只占内容的10-20%，80%以上的token开销被浪费在信息噪声上53% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/855285API
curl https://kongchang.com/api/v1/knowledge/claims/855285MCP
get_claim(id=855285)