Adversarial Attacks and Defenses in Large Language Models: Old and New Threats.
Leo SchwinnDavid DobreStephan GünnemannGauthier GidelPublished in: CoRR (2023)
Keyphrases
- language model
- denial of service attacks
- language modeling
- countermeasures
- denial of service
- dos attacks
- security threats
- document retrieval
- probabilistic model
- n gram
- language modelling
- information retrieval
- speech recognition
- retrieval model
- test collection
- ddos attacks
- query expansion
- intrusion detection system
- data corruption
- statistical language models
- smoothing methods
- cyber attacks
- security measures
- document ranking
- language models for information retrieval
- security vulnerabilities
- attack scenarios
- language model for information retrieval
- security risks
- translation model
- pseudo relevance feedback
- vector space model
- network security
- context sensitive
- query terms
- relevance model
- okapi bm
- ad hoc information retrieval
- security mechanisms
- document length
- web search
- image retrieval
- sql injection
- machine learning