On the Multilingual Ability of Decoder-based Pre-trained Language Models: Finding and Controlling Language-Specific Neurons.
Takeshi KojimaItsuki OkimuraYusuke IwasawaHitomi YanakaYutaka MatsuoPublished in: NAACL-HLT (2024)
Keyphrases
- language specific
- language model
- n gram
- out of vocabulary
- pre trained
- language modeling
- cross lingual
- language independent
- document retrieval
- retrieval model
- probabilistic model
- speech recognition
- machine translation
- natural language
- query expansion
- information retrieval
- neural network
- training data
- query terms
- test collection
- bag of words
- training examples
- translation model
- supervised learning
- text classification
- vector space model
- relevance model
- labor intensive
- video sequences
- text retrieval