Inès de Rosnay

Curious about AI, ML and science

Home
Search
1. Dark Mode

Search

Categories

Artificial Intelligence Networks and Graphs Reflections

Tags

AI Safety Backdoor Cosine Similarity Fine-Tuning Graph Theory LLM Network Analysis PCA QLoRA

Tags

1 page

LLM

Sleeper agents: Training and detecting backdoors in Mistral-7B

Sleeper agents: Training and detecting backdoors in Mistral-7B

© 2026 Inès de Rosnay

Built with Hugo
Theme Stack designed by Jimmy