3 papers
cs.LG2026
Attention-based Classification for Indirect Prompt Injection
Amelia Kawasaki, Stefan Lee
Indirect prompt injection (IPI) attacks, in which adversarial instructions are embedded in external data processed by a language model, pose a significant threat to the integrity o…
cs.CR2025
ShadowLogic: Backdoors in Any Whitebox LLM
Kasimir Schulz, Amelia Kawasaki, Leo Ring
Large language models (LLMs) are widely deployed across various applications, often with safeguards to prevent the generation of harmful or restricted content. However, these safeg…
cs.CR2024
Defending Large Language Models Against Attacks With Residual Stream Activation Analysis
Amelia Kawasaki, Andrew Davis, Houssam Abbas
The widespread adoption of Large Language Models (LLMs), exemplified by OpenAI's ChatGPT, brings to the forefront the imperative to defend against adversarial threats on these mode…