arXiv Science⌕ Search

EXPLORE CONNECTIONS

On Task, Off Track: Benchmarking LLM and Agent Robustness to Near-Target Prompt Injection

Follow the relationships that help you find your next source.

Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.