HyPe: Better Pre-trained Language Model Fine-tuning with Hidden Representation Perturbation.
Hongyi YuanZheng YuanChuanqi TanFei HuangSongfang HuangPublished in: CoRR (2022)
Keyphrases
- language model
- fine tuning
- pre trained
- language modeling
- probabilistic model
- document retrieval
- speech recognition
- n gram
- smoothing methods
- information retrieval
- query expansion
- test collection
- retrieval model
- context sensitive
- translation model
- mixture model
- image representation
- reinforcement learning
- dirichlet prior
- word clouds
- retrieval systems
- relevance model
- training data
- image segmentation