Large pre-trained language models contain human-like biases of what is right and wrong to do.
Patrick SchramowskiCigdem TuranNico AndersenConstantin A. RothkopfKristian KerstingPublished in: Nat. Mach. Intell. (2022)
Keyphrases
- language model
- pre trained
- language modeling
- training data
- n gram
- training examples
- probabilistic model
- speech recognition
- document retrieval
- query expansion
- information retrieval
- retrieval model
- language modelling
- test collection
- control signals
- smoothing methods
- language models for information retrieval
- neural network
- relevance model
- information retrieval systems
- decision trees
- data sets