Targeted attention-head modification, tuned on jailbreak data, lowers attack success rates across models and attacks without fine-tuning.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DETAM: Defending LLMs Against Jailbreak Attacks via Targeted Attention Modification
Targeted attention-head modification, tuned on jailbreak data, lowers attack success rates across models and attacks without fine-tuning.