home
blog
opensource
hardware
research
github
research
Indirect Prompt Injection in Small Language Models
attack taxonomy and size analysis for 0.5B-3B models