Testing AI performance on less frequent aspects of language reveals insensitivity to underlying meaning.
Vittoria DentellaElliot MurphyGary MarcusEvelina LeivadaPublished in: CoRR (2023)
Keyphrases
- natural language
- artificial intelligence
- intended meaning
- formal logic
- artificially intelligent
- computational systems
- programming language
- expert systems
- cognitive architectures
- specification language
- test cases
- language learning
- modeling language
- computer programs
- knowledge representation and reasoning
- semantic structure
- knowledge representation
- language processing
- computational linguistics
- logic programming
- software development
- ai technologies
- relational databases
- case based reasoning
- human cognitive
- machine learning