# data-leakage
標記為「data-leakage」的 17 篇文章
Memory Exfiltration
Techniques for extracting data from AI agent memory systems, including extracting previous conversations, revealing other users' data, and cross-session information leakage.
Privacy Attack Assessment
Test your advanced knowledge of privacy attacks against AI systems including data leakage, PII extraction, differential privacy failures, and inference-time privacy risks with 9 questions.
PII Extraction Techniques
Techniques for extracting personally identifiable information from trained language models including prompt-based extraction, prefix attacks, targeted queries, and real-world examples.
Lab: Output Format Exploitation
Manipulate output formats like JSON, CSV, code blocks, and structured data to extract information that models would normally refuse to provide in natural language.
Weights & Biases Attack Surface
Security analysis of Weights & Biases (W&B/wandb): API key exposure, experiment data leakage, team boundary violations, artifact poisoning, and attack techniques specific to the W&B platform.
Feature Store Access Control
Access control strategies for feature stores: feature-level permissions, cross-team data leakage prevention, PII protection in features, service account management, and implementing least-privilege access for ML feature infrastructure.
KV Cache Poisoning & Exploitation
How KV cache works in transformer inference, cache poisoning across requests in shared deployments, prefix caching attacks, and cross-tenant data leakage.
Structured Output Data Leakage Walkthrough
Walkthrough of using structured output requirements to extract sensitive data embedded in model responses.
Implementing Access Control in RAG Pipelines
Walkthrough for building access control systems in RAG pipelines that enforce document-level permissions, prevent cross-user data leakage, filter retrieved context based on user authorization, and resist retrieval poisoning attacks.
記憶外洩
自 AI 代理記憶系統提取資料之技術，含提取先前對話、揭露其他使用者資料，與跨會話資訊洩漏。
PII Extraction Techniques
Techniques for extracting personally identifiable information from trained language models including prompt-based extraction, prefix attacks, targeted queries, and real-world examples.
實驗室: Output Format 利用ation
Manipulate output formats like JSON, CSV, code blocks, and structured data to extract information that models would normally refuse to provide in natural language.
Weights & Biases 攻擊面
Weights & Biases（W&B／wandb）之安全分析：API 金鑰曝露、實驗資料洩漏、團隊邊界越界、產物投毒，以及 W&B 平台特有之攻擊技術。
Feature Store Access Control
Access control strategies for feature stores: feature-level permissions, cross-team data leakage prevention, PII protection in features, service account management, and implementing least-privilege access for ML feature infrastructure.
KV 快取投毒與利用
KV 快取於 transformer 推論中如何運作、共享部署中的跨請求快取投毒、前綴快取攻擊，以及跨租戶資料洩漏。
Structured Output Data Leakage 導覽
導覽 of using structured output requirements to extract sensitive data embedded in model responses.
Implementing Access Control in RAG Pipelines
導覽 for building access control systems in RAG pipelines that enforce document-level permissions, prevent cross-user data leakage, filter retrieved context based on user authorization, and resist retrieval poisoning attacks.