# pre-training
標記為「pre-training」的 10 篇文章
Training Pipeline Security
Security of the full AI model training pipeline, covering pre-training attacks, fine-tuning and alignment manipulation, architecture-level vulnerabilities, and advanced training-time threats.
Pre-training Attack Surface
Comprehensive overview of pre-training security vulnerabilities including data collection, cleaning, deduplication, and web-scale dataset compromise attack vectors.
Pre-Training Data Attacks
Attacking the pre-training data pipeline including web crawl poisoning and data curation manipulation.
Pre-Training Safety Interventions
Analysis of safety interventions applied during pre-training including data filtering, loss weighting, and curriculum design.
Security Comparison: Pre-training vs Fine-tuning
Comparative analysis of security vulnerabilities, attack surfaces, and defensive strategies across pre-training and fine-tuning phases of language model development.
訓練管線安全
完整 AI 模型訓練管線的安全，涵蓋預訓練攻擊、微調與對齊操控、架構層級漏洞與進階訓練期威脅。
預訓練攻擊面
預訓練安全漏洞的全面概觀，涵蓋資料蒐集、清理、去重，以及 Web 規模資料集入侵攻擊向量。
Pre-訓練 Data 攻擊s
攻擊ing the pre-training data pipeline including web crawl poisoning and data curation manipulation.
Pre-訓練 Safety Interventions
Analysis of safety interventions applied during pre-training including data filtering, loss weighting, and curriculum design.
安全 Comparison: Pre-training vs Fine-tuning
Comparative analysis of security vulnerabilities, attack surfaces, and defensive strategies across pre-training and fine-tuning phases of language model development.