A single learned 3.2-second audio prefix can mute or redirect speech LLMs, and can be trained to selectively mute only targeted genders or languages.
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
-
Universal Acoustic Adversarial Attacks for Flexible Control of Speech-LLMs
A single learned 3.2-second audio prefix can mute or redirect speech LLMs, and can be trained to selectively mute only targeted genders or languages.