PANDAS boosts many-shot jailbreak success rate on open-source LLMs using positive affirmations, negative demonstrations, and Bayesian-optimized topic sampling.
What learning algorithm is in-context learning? investigations with linear models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
PANDAS: Improving Many-shot Jailbreaking via Positive Affirmation, Negative Demonstration, and Adaptive Sampling
PANDAS boosts many-shot jailbreak success rate on open-source LLMs using positive affirmations, negative demonstrations, and Bayesian-optimized topic sampling.