Energy-latency manipulation of multi-modal large language models via verbose samples
Despite the exceptional performance of multi-modal large language models (MLLMs), their deployment requires substantial computational resources. Once malicious users induce high energy consumption and latency time (energy-latency cost), it will exhaust computational resources and harm availability o...
Main Authors: | Gao, K, Gu, J, Bai, Y, Xia, S-T, Torr, P, Liu, W, Li, Z |
---|---|
Format: | Conference item |
Language: | English |
Published: |
2024
|
Similar Items
-
Inducing high energy-latency of large vision-language models with verbose images
by: Gao, K, et al.
Published: (2024) -
The verbosity epidemic.
by: Grais, R, et al.
Published: (2008) -
The Role of Inhibition in Age-Related Off-Topic Verbosity: Not Access but Deletion and Restraint Functions
by: Shufei eYin, et al.
Published: (2016-04-01) -
Review of Verbosity and Description in the Sahifeh Sajjadieh Translation according to Antoine Berman Theory (Case Study: Ansaryan’s Translation)
by: mohamad farhadei, et al.
Published: (2017-09-01) -
New Concepts in Therapeutic Manipulation of HIV-1 Transcription and Latency: Latency Reversal versus Latency Prevention
by: Catherine A. Lewis, et al.
Published: (2023-07-01)