PRISON gives a new roleplay-based benchmark showing current LLMs generate criminal-style traits at high rates but detect them poorly.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
PRISON: Unmasking the Criminal Potential of Large Language Models
PRISON gives a new roleplay-based benchmark showing current LLMs generate criminal-style traits at high rates but detect them poorly.