Predictions from language models for multiple-choice tasks are not robust under variation of scoring methods.
Polina TsvilodubHening WangSharon GroschMichael FrankePublished in: CoRR (2024)
Keyphrases
- language model
- multiple choice
- scoring methods
- language modeling
- n gram
- speech recognition
- document retrieval
- probabilistic model
- language modelling
- query expansion
- statistical language models
- retrieval model
- open ended
- language models for information retrieval
- information retrieval
- vector space model
- test collection
- smoothing methods
- relevance model
- generative model
- statistical language modeling
- document ranking
- learning algorithm
- machine learning