Too Much in Common: Shifting of Embeddings in Transformer Language Models and its Implications.
Daniel BisMaksim PodkorytovXiuwen LiuPublished in: NAACL-HLT (2021)
Keyphrases
- language model
- language modeling
- n gram
- information retrieval
- retrieval model
- speech recognition
- query expansion
- probabilistic model
- document retrieval
- statistical language models
- language modelling
- test collection
- context sensitive
- smoothing methods
- query terms
- ad hoc information retrieval
- vector space model
- vector space
- query specific
- dimensionality reduction
- language models for information retrieval
- pseudo relevance feedback
- relevance model
- term dependencies
- error rate
- distance measure