LogicAttack: Adversarial Attacks for Evaluating Logical Consistency of Natural Language Inference.
Mutsumi NakamuraSantosh MashettyMihir ParmarNeeraj VarshneyChitta BaralPublished in: EMNLP (Findings) (2023)
Keyphrases
- natural language
- representing knowledge
- knowledge representation
- natural language interface
- inference process
- consistency checking
- countermeasures
- natural language processing
- probabilistic inference
- question answering
- multi agent
- machine learning
- information retrieval
- malicious attacks
- semantic interpretation
- natural language understanding
- probabilistic logic
- inference engine
- information extraction
- bayesian networks
- dynamic bayesian networks
- security mechanisms
- constraint networks
- natural language generation
- watermarking scheme
- belief networks
- machine translation
- inference mechanism
- traffic analysis
- knowledge base
- attack detection
- malicious users