Skip to content

ASTPrompter: Preference-Aligned Automated Language Model Red-Teaming to Generate Low-Perplexity Unsafe Prompts.

Amelia Hardy, Houjun Liu, Allie Griffith, Bernard Lange, Duncan Eddy, Mykel J. Kochenderfer

VenueA*EMNLP
Year2025
ProceedingsEMNLP (Findings)

Browse the full EMNLP paper archive.