On the Calibration of Pre-trained Language Models using Mixup Guided by Area Under the Margin and Saliency.
Seoyeon ParkCornelia CarageaPublished in: CoRR (2022)
Keyphrases
- language model
- pre trained
- language modeling
- training data
- document retrieval
- training examples
- probabilistic model
- speech recognition
- information retrieval
- n gram
- query expansion
- test collection
- language modelling
- language models for information retrieval
- statistical language models
- training set
- retrieval model
- visual attention
- control signals
- support vector
- relevance model
- smoothing methods
- data sets
- small number
- semi supervised
- active learning
- machine learning