Unverified50% confidenceFactExact time
Leave-one-family-out cross-validation (LOFO) is a protocol for evaluating generalization to unknown attacks: each time one attack family is completely removed, the model is trained on the remaining families and then tested on the removed family; UNSW-NB15 has 9 attack families, hence 9 folds.
1
Sources
50%
Confidence
Long-term
Relevance
10/22/2026
First Seen
Sources
Related Entities
Related Claims
Unverified间接提示注入攻击中，攻击者只需在RAG会检索到的数据源中植入恶意指令即可操纵模型，传统WAF无法拦截63% similarUnverifiedEcho-like attacks are a typical weakness of RAG architectures: attackers disguise malicious instructions as retrieved result fragments, and the model mistakes them for legitimate instructions and performs operations such as deletion and data exfiltration.56% similarUnverifiedRAG架构引入了间接提示注入攻击面：攻击者只需在RAG检索的数据源中植入恶意指令即可操纵模型55% similarUnverifiedThe introduction of RAG expands the attack surface; attackers only need to have malicious content selected during the retrieval stage to influence outputs, without needing to poison the training data, thereby lowering the threshold for information poisoning.52% similarVerifiedIndirect prompt injection conceals attack instructions within external documents, web pages, or database records retrieved by the model, and is activated when a RAG system introduces them into the context51% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1084106API
curl https://kongchang.com/api/v1/knowledge/claims/1084106MCP
get_claim(id=1084106)