Estimating Large Language Model Capabilities without Labeled Test Data.
Harvey Yiyun FuQinyuan YeAlbert XuXiang RenRobin JiaPublished in: EMNLP (Findings) (2023)
Keyphrases
- test data
- language model
- training data
- language modeling
- training and test data
- training set
- probabilistic model
- document retrieval
- n gram
- test cases
- test set
- speech recognition
- data sets
- language modelling
- context sensitive
- retrieval model
- information retrieval
- test collection
- query expansion
- ad hoc information retrieval
- search based testing
- mixture model
- smoothing methods
- query terms
- supervised learning
- relevance model
- training samples
- information retrieval systems
- translation model
- data points
- prior knowledge
- learning algorithm
- machine learning