Slaves to the Law of Large Numbers: An Asymptotic Equipartition Property for Perplexity in Generative Language Models.
Raghu MudumbaiTyler BellPublished in: CoRR (2024)
Keyphrases
- language model
- language modeling
- probabilistic model
- n gram
- generative model
- document retrieval
- information retrieval
- language modelling
- word error rate
- speech recognition
- retrieval model
- query expansion
- test collection
- context sensitive
- smoothing methods
- language modeling framework
- statistical language models
- ad hoc information retrieval
- relevance model
- translation model
- feature selection
- query terms
- machine learning
- co occurrence
- vector space model
- term dependencies
- text retrieval
- web pages
- information retrieval systems